{"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/the-narrativeqa-reading-comprehension","title":"The NarrativeQA Reading Comprehension Challenge","arxiv_id":"1712.07040","date":"2017-12-19","proceeding":"TACL 2018 1","authors":["Tomáš Kočiský","Jonathan Schwarz","Phil Blunsom","Chris Dyer","Karl Moritz Hermann","Gábor Melis","Edward Grefenstette"],"abstract":"Reading comprehension (RC)---in contrast to information retrieval---requires\nintegrating information and reasoning about events, entities, and their\nrelations across a full document. Question answering is conventionally used to\nassess RC ability, in both artificial agents and children learning to read.\nHowever, existing RC datasets and tasks are dominated by questions that can be\nsolved by selecting answers using superficial information (e.g., local context\nsimilarity or global term frequency); they thus fail to test for the essential\nintegrative aspect of RC. To encourage progress on deeper comprehension of\nlanguage, we present a new dataset and set of tasks in which the reader must\nanswer questions about stories by reading entire books or movie scripts. These\ntasks are designed so that successfully answering their questions requires\nunderstanding the underlying narrative rather than relying on shallow pattern\nmatching or salience. We show that although humans solve the tasks easily,\nstandard RC models struggle on the tasks presented here. We provide an analysis\nof the dataset and the challenges it presents.","url_abs":"http://arxiv.org/abs/1712.07040v1","url_pdf":"http://arxiv.org/pdf/1712.07040v1.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":"the-narrativeqa-reading-comprehension","repo_url":"https://github.com/deepmind/narrativeqa","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"unanswered"}},{"paper_slug":"the-narrativeqa-reading-comprehension","repo_url":"https://github.com/google-deepmind/narrativeqa","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"unanswered"}}],"tasks":[{"task_slug":"information-retrieval","task_name":"Information Retrieval"},{"task_slug":"question-answering","task_name":"Question Answering"},{"task_slug":"reading-comprehension","task_name":"Reading Comprehension"},{"task_slug":"retrieval","task_name":"Retrieval"}],"methods":[],"datasets_introduced":[{"slug":"narrativeqa","name":"NarrativeQA","full_name":"NarrativeQA"}],"methods_introduced":[],"results":[{"leaderboard":"/sota/question-answering-on-narrativeqa","task":"Question Answering","dataset":"NarrativeQA","model":"Oracle IR Models","rank_in_archive_order":10,"of":10,"metrics":{"BLEU-1":"54.60/55.55","BLEU-4":"26.71/27.78"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1712.07040","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}