{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/multi-perspective-context-matching-for","title":"Multi-Perspective Context Matching for Machine Comprehension","arxiv_id":"1612.04211","date":"2016-12-13","proceeding":null,"authors":["Zhiguo Wang","Haitao Mi","Wael Hamza","Radu Florian"],"abstract":"Previous machine comprehension (MC) datasets are either too small to train\nend-to-end deep learning models, or not difficult enough to evaluate the\nability of current MC techniques. The newly released SQuAD dataset alleviates\nthese limitations, and gives us a chance to develop more realistic MC models.\nBased on this dataset, we propose a Multi-Perspective Context Matching (MPCM)\nmodel, which is an end-to-end system that directly predicts the answer\nbeginning and ending points in a passage. Our model first adjusts each\nword-embedding vector in the passage by multiplying a relevancy weight computed\nagainst the question. Then, we encode the question and weighted passage by\nusing bi-directional LSTMs. For each point in the passage, our model matches\nthe context of this point against the encoded question from multiple\nperspectives and produces a matching vector. Given those matched vectors, we\nemploy another bi-directional LSTM to aggregate all the information and predict\nthe beginning and ending points. Experimental result on the test set of SQuAD\nshows that our model achieves a competitive result on the leaderboard.","url_abs":"http://arxiv.org/abs/1612.04211v1","url_pdf":"http://arxiv.org/pdf/1612.04211v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"multi-perspective-context-matching-for","repo_url":"https://github.com/bloomsburyai/question-generation","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"unanswered"}}],"tasks":[{"task_slug":"question-answering","task_name":"Question Answering"},{"task_slug":"reading-comprehension","task_name":"Reading Comprehension"}],"methods":[{"method_slug":"lstm","method_name":"LSTM"},{"method_slug":"sigmoid-activation","method_name":"Sigmoid Activation"},{"method_slug":"tanh-activation","method_name":"Tanh Activation"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/open-domain-question-answering-on-squad11","task":"Open-Domain Question Answering","dataset":"SQuAD1.1","model":"MPCM","rank_in_archive_order":3,"of":3,"metrics":{"EM":"65.5"},"uses_additional_data":false},{"leaderboard":"/sota/question-answering-on-squad11","task":"Question Answering","dataset":"SQuAD1.1","model":"Multi-Perspective Matching (ensemble)","rank_in_archive_order":133,"of":213,"metrics":{"EM":"73.765","F1":"81.257"},"uses_additional_data":false},{"leaderboard":"/sota/question-answering-on-squad11","task":"Question Answering","dataset":"SQuAD1.1","model":"Multi-Perspective Matching (single model)","rank_in_archive_order":162,"of":213,"metrics":{"EM":"70.387","F1":"78.784"},"uses_additional_data":false},{"leaderboard":"/sota/question-answering-on-squad11-dev","task":"Question Answering","dataset":"SQuAD1.1 dev","model":"MPCM","rank_in_archive_order":44,"of":55,"metrics":{"EM":"66.1","F1":"75.8"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1612.04211","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}