{"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/neural-multi-step-reasoning-for-question","title":"Neural Multi-Step Reasoning for Question Answering on Semi-Structured Tables","arxiv_id":"1702.06589","date":"2017-02-21","proceeding":null,"authors":["Till Haug","Octavian-Eugen Ganea","Paulina Grnarova"],"abstract":"Advances in natural language processing tasks have gained momentum in recent\nyears due to the increasingly popular neural network methods. In this paper, we\nexplore deep learning techniques for answering multi-step reasoning questions\nthat operate on semi-structured tables. Challenges here arise from the level of\nlogical compositionality expressed by questions, as well as the domain\nopenness. Our approach is weakly supervised, trained on question-answer-table\ntriples without requiring intermediate strong supervision. It performs two\nphases: first, machine understandable logical forms (programs) are generated\nfrom natural language questions following the work of [Pasupat and Liang,\n2015]. Second, paraphrases of logical forms and questions are embedded in a\njointly learned vector space using word and character convolutional neural\nnetworks. A neural scoring function is further used to rank and retrieve the\nmost probable logical form (interpretation) of a question. Our best single\nmodel achieves 34.8% accuracy on the WikiTableQuestions dataset, while the best\nensemble of our models pushes the state-of-the-art score on this task to 38.7%,\nthus slightly surpassing both the engineered feature scoring baseline, as well\nas the Neural Programmer model of [Neelakantan et al., 2016].","url_abs":"http://arxiv.org/abs/1702.06589v2","url_pdf":"http://arxiv.org/pdf/1702.06589v2.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":"neural-multi-step-reasoning-for-question","repo_url":"https://github.com/dalab/neural_qa","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"tf","reach":null}],"tasks":[{"task_slug":"question-answering","task_name":"Question Answering"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1702.06589","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}