{"id":215081,"metadata":{"_oai":{"id":"oai:ipsj.ixsq.nii.ac.jp:00215081","sets":["6504:10735:10808"]},"path":["10808"],"owner":"44499","recid":"215081","title":["汎用言語モデルにco-matching modelを追加した多肢選択機械読解に関する一検討"],"pubdate":{"attribute_name":"公開日","attribute_value":"2021-03-04"},"_buckets":{"deposit":"7b423fc8-6208-45b2-b413-1565ae86d8de"},"_deposit":{"id":"215081","pid":{"type":"depid","value":"215081","revision_id":0},"owners":[44499],"status":"published","created_by":44499},"item_title":"汎用言語モデルにco-matching modelを追加した多肢選択機械読解に関する一検討","author_link":["553592","553591"],"item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"汎用言語モデルにco-matching modelを追加した多肢選択機械読解に関する一検討"}]},"item_keyword":{"attribute_name":"キーワード","attribute_value_mlt":[{"subitem_subject":"人工知能と認知科学","subitem_subject_scheme":"Other"}]},"item_type_id":"22","publish_date":"2021-03-04","item_language":{"attribute_name":"言語","attribute_value_mlt":[{"subitem_language":"jpn"}]},"item_22_text_3":{"attribute_name":"著者所属","attribute_value_mlt":[{"subitem_text_value":"東京都市大"},{"subitem_text_value":"東京都市大"}]},"item_publisher":{"attribute_name":"出版者","attribute_value_mlt":[{"subitem_publisher":"情報処理学会","subitem_publisher_language":"ja"}]},"publish_status":"0","weko_shared_id":-1,"item_file_price":{"attribute_name":"Billing file","attribute_type":"file","attribute_value_mlt":[{"url":{"url":"https://ipsj.ixsq.nii.ac.jp/record/215081/files/IPSJ-Z83-4S-04.pdf","label":"IPSJ-Z83-4S-04.pdf"},"date":[{"dateType":"Available","dateValue":"2021-12-28"}],"format":"application/pdf","filename":"IPSJ-Z83-4S-04.pdf","filesize":[{"value":"810.7 kB"}],"mimetype":"application/pdf","accessrole":"open_date","version_id":"549973ac-d7c7-4df3-a4fd-d9f5a2a46a13","displaytype":"detail","licensetype":"license_note","license_note":"Copyright (c) 2021 by the Information Processing Society of Japan"}]},"item_22_creator_5":{"attribute_name":"著者名","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"藤田, 和成"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"荒井, 秀一"}],"nameIdentifiers":[{}]}]},"item_resource_type":{"attribute_name":"資源タイプ","attribute_value_mlt":[{"resourceuri":"http://purl.org/coar/resource_type/c_5794","resourcetype":"conference paper"}]},"item_22_source_id_9":{"attribute_name":"書誌レコードID","attribute_value_mlt":[{"subitem_source_identifier":"AN00349328","subitem_source_identifier_type":"NCID"}]},"item_22_description_7":{"attribute_name":"論文抄録","attribute_value_mlt":[{"subitem_description":"対話ベースの多肢選択機械読解のDreamでは，XLNetなどの汎用言語モデルを用いた手法が多く提案されている．その多くの手法では，質問文と選択肢を1つに連結して本文とのマッチングを行なうため，質問文を無視し選択肢の単語に影響を受け誤答することがある．本稿では本文-質問文と本文-選択肢の各々の重要箇所の抽出を独立して行うことが重要と考え，Co-matching modelを汎用言語モデルに追加し，質問文と選択肢を明示的に2つのシーケンスに分けて本文とのマッチングを行うことを考えた．一例として汎用言語モデルにXLNetを用いた場合，3.8pointの精度向上を確認することができた．","subitem_description_type":"Other"}]},"item_22_biblio_info_10":{"attribute_name":"書誌情報","attribute_value_mlt":[{"bibliographicPageEnd":"566","bibliographic_titles":[{"bibliographic_title":"第83回全国大会講演論文集"}],"bibliographicPageStart":"565","bibliographicIssueDates":{"bibliographicIssueDate":"2021-03-04","bibliographicIssueDateType":"Issued"},"bibliographicIssueNumber":"1","bibliographicVolumeNumber":"2021"}]},"relation_version_is_last":true,"weko_creator_id":"44499"},"updated":"2025-01-19T16:19:50.405689+00:00","created":"2025-01-19T01:15:51.813178+00:00","links":{}}