{"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/variational-reasoning-for-question-answering","title":"Variational Reasoning for Question Answering with Knowledge Graph","arxiv_id":"1709.04071","date":"2017-09-12","proceeding":null,"authors":["Yuyu Zhang","Hanjun Dai","Zornitsa Kozareva","Alexander J. Smola","Le Song"],"abstract":"Knowledge graph (KG) is known to be helpful for the task of question\nanswering (QA), since it provides well-structured relational information\nbetween entities, and allows one to further infer indirect facts. However, it\nis challenging to build QA systems which can learn to reason over knowledge\ngraphs based on question-answer pairs alone. First, when people ask questions,\ntheir expressions are noisy (for example, typos in texts, or variations in\npronunciations), which is non-trivial for the QA system to match those\nmentioned entities to the knowledge graph. Second, many questions require\nmulti-hop logic reasoning over the knowledge graph to retrieve the answers. To\naddress these challenges, we propose a novel and unified deep learning\narchitecture, and an end-to-end variational learning algorithm which can handle\nnoise in questions, and learn multi-hop reasoning simultaneously. Our method\nachieves state-of-the-art performance on a recent benchmark dataset in the\nliterature. We also derive a series of new benchmark datasets, including\nquestions for multi-hop reasoning, questions paraphrased by neural translation\nmodel, and questions in human voice. Our method yields very promising results\non all these challenging datasets.","url_abs":"http://arxiv.org/abs/1709.04071v5","url_pdf":"http://arxiv.org/pdf/1709.04071v5.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":"variational-reasoning-for-question-answering","repo_url":"https://github.com/yuyuz/Variational-Reasoning-Networks","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"none","reach":{"status":"unanswered"}}],"tasks":[{"task_slug":"knowledge-graphs","task_name":"Knowledge Graphs"},{"task_slug":"question-answering","task_name":"Question Answering"},{"task_slug":"translation","task_name":"Translation"}],"methods":[],"datasets_introduced":[{"slug":"metaqa","name":"MetaQA","full_name":"MoviE Text Audio QA"}],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1709.04071","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}