{"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/mapping-to-declarative-knowledge-for-word","title":"Mapping to Declarative Knowledge for Word Problem Solving","arxiv_id":"1712.09391","date":"2017-12-26","proceeding":"TACL 2018 1","authors":["Subhro Roy","Dan Roth"],"abstract":"Math word problems form a natural abstraction to a range of quantitative\nreasoning problems, such as understanding financial news, sports results, and\ncasualties of war. Solving such problems requires the understanding of several\nmathematical concepts such as dimensional analysis, subset relationships, etc.\nIn this paper, we develop declarative rules which govern the translation of\nnatural language description of these concepts to math expressions. We then\npresent a framework for incorporating such declarative knowledge into word\nproblem solving. Our method learns to map arithmetic word problem text to math\nexpressions, by learning to select the relevant declarative knowledge for each\noperation of the solution expression. This provides a way to handle multiple\nconcepts in the same problem while, at the same time, support interpretability\nof the answer expression. Our method models the mapping to declarative\nknowledge as a latent variable, thus removing the need for expensive\nannotations. Experimental evaluation suggests that our domain knowledge based\nsolver outperforms all other systems, and that it generalizes better in the\nrealistic case where the training data it is exposed to is biased in a\ndifferent way than the test data.","url_abs":"http://arxiv.org/abs/1712.09391v1","url_pdf":"http://arxiv.org/pdf/1712.09391v1.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":"mapping-to-declarative-knowledge-for-word","repo_url":"https://github.com/CogComp/arithmetic","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"none","reach":{"status":"unanswered"}}],"tasks":[{"task_slug":"math","task_name":"Math"},{"task_slug":"translation","task_name":"Translation"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1712.09391","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}