{"created":"2023-05-15T08:43:59.306735+00:00","id":9170,"links":{},"metadata":{"_buckets":{"deposit":"dc791ee8-a4eb-4ecc-906f-255eef674eee"},"_deposit":{"created_by":13,"id":"9170","owners":[13],"pid":{"revision_id":0,"type":"depid","value":"9170"},"status":"published"},"_oai":{"id":"oai:uec.repo.nii.ac.jp:00009170","sets":["6"]},"author_link":["25073","25074","25075","25076","25077"],"control_number":"9170","item_10001_biblio_info_7":{"attribute_name":"書誌情報","attribute_value_mlt":[{"bibliographicIssueDates":{"bibliographicIssueDate":"2011-11-01","bibliographicIssueDateType":"Issued"},"bibliographicIssueNumber":"11","bibliographicPageEnd":"1761","bibliographicPageStart":"1751","bibliographicVolumeNumber":"J94-D","bibliographic_titles":[{"bibliographic_title":"電子情報通信学会論文誌. D, 情報・システム","bibliographic_titleLang":"ja"}]}]},"item_10001_description_5":{"attribute_name":"抄録","attribute_value_mlt":[{"subitem_description":"インターネットの普及に伴い，Web上には商品やサービス（対象物）に対する多くの評判情報が蓄積されている．しかし，誰でも発信できるが故に，情報は膨大となり，それら全てに目を通すことは利用者にとって多大な負担となる．そのような背景から，レビューから意見を自動的に抽出する研究が盛んに行われており，意見を<対象物，評価視点（属性），評価値>の三つ組と捉え抽出する研究も行われている．しかしながら多くの研究は，評価視点や評価値の抽出に辞書を用いており，ジャンルごとに必要となる辞書の構築のためのコストは小さいとはいえない．また，単に辞書を用いてマッチングを行っただけでは，精度が上がらないのが現状である．そこで本論文では，教師あり学習を用いて，レビューサイトから意見を抽出する手法を提案する．提案手法は，従来の手法のように大規模な辞書をあらかじめ用意する必要がないため，コストを大幅に抑えることが可能である．実験の結果，辞書をあらかじめ用意しない既存手法と比較して，最大で適合率は約26%，再現率は約47%向上した．また，既存研究では個別の辞書を必要とするような他ジャンルに対して本手法を適用した結果，ほぼ変わらない精度で抽出することができ，他ジャンルへの適用の可能性を示すことができた．","subitem_description_type":"Abstract"},{"subitem_description":"In recent years, most research have been done on automatic extraction of opinions (object, attribute, value) from a review article on the Web. However, most of them use a dictionary, which need a huge cost, to extract attributes and its values for each genre of products. In this paper, we proposed a method using supervised learning to extract opinions without using a large dictionary. The experimental results show that our method could generate up to 26% of precision and 47% of recall higher than conventional methods. We also applied our method to a different genre and received almost the same result.","subitem_description_type":"Abstract"}]},"item_10001_publisher_8":{"attribute_name":"出版者","attribute_value_mlt":[{"subitem_publisher":"電子情報通信学会"}]},"item_10001_relation_17":{"attribute_name":"関連サイト","attribute_value_mlt":[{"subitem_relation_type_id":{"subitem_relation_type_id_text":"http://search.ieice.org/index.html","subitem_relation_type_select":"URI"}}]},"item_10001_rights_15":{"attribute_name":"権利","attribute_value_mlt":[{"subitem_rights":"Copyright © 2011 IEICE"}]},"item_10001_source_id_9":{"attribute_name":"ISSN","attribute_value_mlt":[{"subitem_source_identifier":"1881-0225","subitem_source_identifier_type":"ISSN"}]},"item_10001_version_type_20":{"attribute_name":"著者版フラグ","attribute_value_mlt":[{"subitem_version_resource":"http://purl.org/coar/version/c_970fb48d4fbd8a85","subitem_version_type":"VoR"}]},"item_creator":{"attribute_name":"著者","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"田中, 俊行","creatorNameLang":"ja"},{"creatorName":"タナカ, トシユキ","creatorNameLang":"ja-Kana"},{"creatorName":"TANAKA, Toshiyuki","creatorNameLang":"en"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"グエン, ミン テイ","creatorNameLang":"ja"},{"creatorName":"NGUYEN, The-Minh","creatorNameLang":"en"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"中川, 博之","creatorNameLang":"ja"},{"creatorName":"ナカガワ, ヒロユキ","creatorNameLang":"ja-Kana"},{"creatorName":"NAKAGAWA, Hiroyuki","creatorNameLang":"en"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"田原, 康之","creatorNameLang":"ja"},{"creatorName":"タハラ, ヤスユキ","creatorNameLang":"ja-Kana"},{"creatorName":"TAHARA, Yasuyuki","creatorNameLang":"en"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"大須賀, 昭彦","creatorNameLang":"ja"},{"creatorName":"オオスガ, アキヒコ","creatorNameLang":"ja-Kana"},{"creatorName":"OHSUGA, Akihiko","creatorNameLang":"en"}],"nameIdentifiers":[{}]}]},"item_files":{"attribute_name":"ファイル情報","attribute_type":"file","attribute_value_mlt":[{"accessrole":"open_date","date":[{"dateType":"Available","dateValue":"2019-05-08"}],"displaytype":"detail","filename":"j94-d_11_1751.pdf","filesize":[{"value":"728.9 kB"}],"format":"application/pdf","licensetype":"license_note","mimetype":"application/pdf","url":{"label":"j94-d_11_1751","url":"https://uec.repo.nii.ac.jp/record/9170/files/j94-d_11_1751.pdf"},"version_id":"9940d5a6-9e42-43e8-8792-587f89383c3e"}]},"item_keyword":{"attribute_name":"キーワード","attribute_value_mlt":[{"subitem_subject":"意見抽出","subitem_subject_scheme":"Other"},{"subitem_subject":"評判情報","subitem_subject_scheme":"Other"},{"subitem_subject":"機械学習","subitem_subject_scheme":"Other"},{"subitem_subject":"opinion extraction","subitem_subject_language":"en","subitem_subject_scheme":"Other"},{"subitem_subject":"reputation","subitem_subject_language":"en","subitem_subject_scheme":"Other"},{"subitem_subject":"machine learning","subitem_subject_language":"en","subitem_subject_scheme":"Other"}]},"item_language":{"attribute_name":"言語","attribute_value_mlt":[{"subitem_language":"jpn"}]},"item_resource_type":{"attribute_name":"資源タイプ","attribute_value_mlt":[{"resourcetype":"journal article","resourceuri":"http://purl.org/coar/resource_type/c_6501"}]},"item_title":"評判分析システムのための教師あり学習を用いた意見抽出","item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"評判分析システムのための教師あり学習を用いた意見抽出","subitem_title_language":"ja"},{"subitem_title":"Opinion Extraction Using Supervised Learning for Reputation Analysis System","subitem_title_language":"en"}]},"item_type_id":"10001","owner":"13","path":["6"],"pubdate":{"attribute_name":"PubDate","attribute_value":"2019-05-08"},"publish_date":"2019-05-08","publish_status":"0","recid":"9170","relation_version_is_last":true,"title":["評判分析システムのための教師あり学習を用いた意見抽出"],"weko_creator_id":"13","weko_shared_id":-1},"updated":"2024-03-04T01:56:31.272836+00:00"}