{"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/what-makes-reading-comprehension-questions","title":"What Makes Reading Comprehension Questions Easier?","arxiv_id":"1808.09384","date":"2018-08-28","proceeding":"EMNLP 2018 10","authors":["Saku Sugawara","Kentaro Inui","Satoshi Sekine","Akiko Aizawa"],"abstract":"A challenge in creating a dataset for machine reading comprehension (MRC) is\nto collect questions that require a sophisticated understanding of language to\nanswer beyond using superficial cues. In this work, we investigate what makes\nquestions easier across recent 12 MRC datasets with three question styles\n(answer extraction, description, and multiple choice). We propose to employ\nsimple heuristics to split each dataset into easy and hard subsets and examine\nthe performance of two baseline models for each of the subsets. We then\nmanually annotate questions sampled from each subset with both validity and\nrequisite reasoning skills to investigate which skills explain the difference\nbetween easy and hard questions. From this study, we observed that (i) the\nbaseline performances for the hard subsets remarkably degrade compared to those\nof entire datasets, (ii) hard questions require knowledge inference and\nmultiple-sentence reasoning in comparison with easy questions, and (iii)\nmultiple-choice questions tend to require a broader range of reasoning skills\nthan answer extraction and description questions. These results suggest that\none might overestimate recent advances in MRC.","url_abs":"http://arxiv.org/abs/1808.09384v1","url_pdf":"http://arxiv.org/pdf/1808.09384v1.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":"what-makes-reading-comprehension-questions","repo_url":"https://github.com/Alab-NII/mrc-heuristics","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"none","reach":{"status":"ok"}}],"tasks":[{"task_slug":"machine-reading-comprehension","task_name":"Machine Reading Comprehension"},{"task_slug":"multiple-choice","task_name":"Multiple-choice"},{"task_slug":"reading-comprehension","task_name":"Reading Comprehension"},{"task_slug":"sentence","task_name":"Sentence"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1808.09384","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}