{"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-granularity-hierarchical-attention","title":"Multi-granularity hierarchical attention fusion networks for reading comprehension and question answering","arxiv_id":"1811.11934","date":"2018-11-29","proceeding":"ACL 2018 7","authors":["Wei Wang","Ming Yan","Chen Wu"],"abstract":"This paper describes a novel hierarchical attention network for reading\ncomprehension style question answering, which aims to answer questions for a\ngiven narrative paragraph. In the proposed method, attention and fusion are\nconducted horizontally and vertically across layers at different levels of\ngranularity between question and paragraph. Specifically, it first encode the\nquestion and paragraph with fine-grained language embeddings, to better capture\nthe respective representations at semantic level. Then it proposes a\nmulti-granularity fusion approach to fully fuse information from both global\nand attended representations. Finally, it introduces a hierarchical attention\nnetwork to focuses on the answer span progressively with multi-level\nsoftalignment. Extensive experiments on the large-scale SQuAD and TriviaQA\ndatasets validate the effectiveness of the proposed method. At the time of\nwriting the paper (Jan. 12th 2018), our model achieves the first position on\nthe SQuAD leaderboard for both single and ensemble models. We also achieves\nstate-of-the-art results on TriviaQA, AddSent and AddOne-Sent datasets.","url_abs":"http://arxiv.org/abs/1811.11934v1","url_pdf":"http://arxiv.org/pdf/1811.11934v1.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-granularity-hierarchical-attention","repo_url":"https://github.com/alibaba/AliceMind/tree/main/StructBERT","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"question-answering","task_name":"Question Answering"},{"task_slug":"reading-comprehension","task_name":"Reading Comprehension"},{"task_slug":"triviaqa","task_name":"TriviaQA"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1811.11934","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}