{"id":19962,"updated":"2025-01-22T21:58:05.384205+00:00","links":{},"created":"2025-01-18T22:52:17.116753+00:00","metadata":{"_oai":{"id":"oai:ipsj.ixsq.nii.ac.jp:00019962","sets":["1164:1165:1209:1211"]},"path":["1211"],"owner":"1","recid":"19962","title":["単語の重要度に基づくテキストの要約"],"pubdate":{"attribute_name":"公開日","attribute_value":"1998-05-14"},"_buckets":{"deposit":"2d6c0af6-f7e9-47f7-a7bc-e40c7736e9c3"},"_deposit":{"id":"19962","pid":{"type":"depid","value":"19962","revision_id":0},"owners":[1],"status":"published","created_by":1},"item_title":"単語の重要度に基づくテキストの要約","author_link":["0","0"],"item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"単語の重要度に基づくテキストの要約"},{"subitem_title":"Text Summarization based on Word Importance","subitem_title_language":"en"}]},"item_type_id":"4","publish_date":"1998-05-14","item_4_text_3":{"attribute_name":"著者所属","attribute_value_mlt":[{"subitem_text_value":"NTTデータ通信（株）情報科学研究所"},{"subitem_text_value":"NTTデータ通信（株）情報科学研究所"}]},"item_4_text_4":{"attribute_name":"著者所属(英)","attribute_value_mlt":[{"subitem_text_value":"Laboratory for Information Technology NTT DATA Corporation","subitem_text_language":"en"},{"subitem_text_value":"Laboratory for Information Technology NTT DATA Corporation","subitem_text_language":"en"}]},"item_language":{"attribute_name":"言語","attribute_value_mlt":[{"subitem_language":"jpn"}]},"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/19962/files/IPSJ-DBS98115006.pdf"},"date":[{"dateType":"Available","dateValue":"2000-05-14"}],"format":"application/pdf","billing":["billing_file"],"filename":"IPSJ-DBS98115006.pdf","filesize":[{"value":"660.9 kB"}],"mimetype":"application/pdf","priceinfo":[{"tax":["include_tax"],"price":"660","billingrole":"5"},{"tax":["include_tax"],"price":"330","billingrole":"6"},{"tax":["include_tax"],"price":"0","billingrole":"13"},{"tax":["include_tax"],"price":"0","billingrole":"44"}],"accessrole":"open_date","version_id":"a62175e0-e8b7-4917-94ef-a4ec3b33ebb1","displaytype":"detail","licensetype":"license_note","license_note":"Copyright (c) 1998 by the Information Processing Society of Japan"}]},"item_4_creator_5":{"attribute_name":"著者名","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"平尾, 努"},{"creatorName":"木谷強"}],"nameIdentifiers":[{}]}]},"item_4_creator_6":{"attribute_name":"著者名(英)","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"Tsutomu, Hirao","creatorNameLang":"en"},{"creatorName":"Tsuyoshi, Kitani","creatorNameLang":"en"}],"nameIdentifiers":[{}]}]},"item_4_source_id_9":{"attribute_name":"書誌レコードID","attribute_value_mlt":[{"subitem_source_identifier":"AN10112482","subitem_source_identifier_type":"NCID"}]},"item_4_textarea_12":{"attribute_name":"Notice","attribute_value_mlt":[{"subitem_textarea_value":"SIG Technical Reports are nonrefereed and hence may later appear in any journals, conferences, symposia, etc."}]},"item_resource_type":{"attribute_name":"資源タイプ","attribute_value_mlt":[{"resourceuri":"http://purl.org/coar/resource_type/c_18gh","resourcetype":"technical report"}]},"item_4_description_7":{"attribute_name":"論文抄録","attribute_value_mlt":[{"subitem_description":"本稿では，単語の範囲内重要度に基づき文の重要度を決定し，要約文を抽出する方法について述べる．提案する手法では，単に重要度の高い文を要約文とするのではなく，隣接する文の重要度の変化に着目して要約文を抽出する．新聞記事を対象として実験を行ない，人手により抽出した正解データから適合率・再現率を求める方法と，提案手法により抽出した要約文に対するアンケート調査とによって提案手法を評価した．評価の結果，提案手法により，人間が最重要と考える文を平均63％の精度で抽出することができた．また，提案手法を発展させることにより，精度の向上が期待できることもわかった．","subitem_description_type":"Other"}]},"item_4_description_8":{"attribute_name":"論文抄録(英)","attribute_value_mlt":[{"subitem_description":"In this paper, we describe a method of determining the importance of sentences based on word importance within a text region and extracting summary sentences. The summary sentences are not simply chosen by the importance, but are determined by the change of importance between adjacent sensences. This proposed technique was evaluated using newspaper articles. Two different evaluation methods: recall and precision rates by comparing the extracted summaries with manually selected ones, and human assessments to the extracted summaries. The proposed method extracted sentences that the human assessor judged as the most important on an average of 63%. We should be able to achieve a better accuracy when the proposal technique is futher improved.","subitem_description_type":"Other"}]},"item_4_biblio_info_10":{"attribute_name":"書誌情報","attribute_value_mlt":[{"bibliographicPageEnd":"48","bibliographic_titles":[{"bibliographic_title":"情報処理学会研究報告データベースシステム（DBS）"}],"bibliographicPageStart":"41","bibliographicIssueDates":{"bibliographicIssueDate":"1998-05-14","bibliographicIssueDateType":"Issued"},"bibliographicIssueNumber":"34(1998-DBS-115)","bibliographicVolumeNumber":"1998"}]},"relation_version_is_last":true,"weko_creator_id":"1"}}