{"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/a-deep-architecture-for-semantic-matching","title":"A Deep Architecture for Semantic Matching with Multiple Positional Sentence Representations","arxiv_id":"1511.08277","date":"2015-11-26","proceeding":null,"authors":["Shengxian Wan","Yanyan Lan","Jiafeng Guo","Jun Xu","Liang Pang","Xue-Qi Cheng"],"abstract":"Matching natural language sentences is central for many applications such as\ninformation retrieval and question answering. Existing deep models rely on a\nsingle sentence representation or multiple granularity representations for\nmatching. However, such methods cannot well capture the contextualized local\ninformation in the matching process. To tackle this problem, we present a new\ndeep architecture to match two sentences with multiple positional sentence\nrepresentations. Specifically, each positional sentence representation is a\nsentence representation at this position, generated by a bidirectional long\nshort term memory (Bi-LSTM). The matching score is finally produced by\naggregating interactions between these different positional sentence\nrepresentations, through $k$-Max pooling and a multi-layer perceptron. Our\nmodel has several advantages: (1) By using Bi-LSTM, rich context of the whole\nsentence is leveraged to capture the contextualized local information in each\npositional sentence representation; (2) By matching with multiple positional\nsentence representations, it is flexible to aggregate different important\ncontextualized local information in a sentence to support the matching; (3)\nExperiments on different tasks such as question answering and sentence\ncompletion demonstrate the superiority of our model.","url_abs":"http://arxiv.org/abs/1511.08277v1","url_pdf":"http://arxiv.org/pdf/1511.08277v1.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":"a-deep-architecture-for-semantic-matching","repo_url":"https://github.com/jastfkjg/semantic-matching","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok"}}],"tasks":[{"task_slug":"information-retrieval","task_name":"Information Retrieval"},{"task_slug":"question-answering","task_name":"Question Answering"},{"task_slug":"retrieval","task_name":"Retrieval"},{"task_slug":"sentence","task_name":"Sentence"},{"task_slug":"sentence-completion","task_name":"Sentence Completion"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":"https://app.syntology.ai/?focus=1511.08277","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}