{"metadata":{"_oai":{"id":"oai:ipsj.ixsq.nii.ac.jp:00011605","sets":["581:664:671"]},"path":["671"],"owner":"1","recid":"11605","title":["トランスダクティブ・ブースティング法によるテキスト分類"],"pubdate":{"attribute_name":"公開日","attribute_value":"2002-06-15"},"_buckets":{"deposit":"2dbb43e3-e218-46ff-b6d8-1a4ff9ae2656"},"_deposit":{"id":"11605","pid":{"type":"depid","value":"11605","revision_id":0},"owners":[1],"status":"published","created_by":1},"item_title":"トランスダクティブ・ブースティング法によるテキスト分類","author_link":["0","0"],"item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"トランスダクティブ・ブースティング法によるテキスト分類"},{"subitem_title":"Text Categorization Using a Transductive Boosting Method","subitem_title_language":"en"}]},"item_keyword":{"attribute_name":"キーワード","attribute_value_mlt":[{"subitem_subject":"論文","subitem_subject_scheme":"Other"}]},"item_type_id":"2","publish_date":"2002-06-15","item_2_text_3":{"attribute_name":"著者所属","attribute_value_mlt":[{"subitem_text_value":"NTTコミュニケーション科学基礎研究所"},{"subitem_text_value":"ATR人間情報科学研究所"}]},"item_2_text_4":{"attribute_name":"著者所属(英)","attribute_value_mlt":[{"subitem_text_value":"NTT Communication Science Laboratories","subitem_text_language":"en"},{"subitem_text_value":"ATR Human Information Science Laboratories","subitem_text_language":"en"}]},"item_language":{"attribute_name":"言語","attribute_value_mlt":[{"subitem_language":"jpn"}]},"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/11605/files/IPSJ-JNL4306028.pdf"},"date":[{"dateType":"Available","dateValue":"2004-06-15"}],"format":"application/pdf","billing":["billing_file"],"filename":"IPSJ-JNL4306028.pdf","filesize":[{"value":"206.9 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":"8"},{"tax":["include_tax"],"price":"0","billingrole":"44"}],"accessrole":"open_date","version_id":"6cb2bbe4-9c14-42e5-a2b8-d649550d0abd","displaytype":"detail","licensetype":"license_note","license_note":"Copyright (c) 2002 by the Information Processing Society of Japan"}]},"item_2_creator_5":{"attribute_name":"著者名","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"平, 博順"},{"creatorName":"春野, 雅彦"}],"nameIdentifiers":[{}]}]},"item_2_creator_6":{"attribute_name":"著者名(英)","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"Hirotoshi, Taira","creatorNameLang":"en"},{"creatorName":"Masahiko, Haruno","creatorNameLang":"en"}],"nameIdentifiers":[{}]}]},"item_2_source_id_9":{"attribute_name":"書誌レコードID","attribute_value_mlt":[{"subitem_source_identifier":"AN00116647","subitem_source_identifier_type":"NCID"}]},"item_resource_type":{"attribute_name":"資源タイプ","attribute_value_mlt":[{"resourceuri":"http://purl.org/coar/resource_type/c_6501","resourcetype":"journal article"}]},"item_2_source_id_11":{"attribute_name":"ISSN","attribute_value_mlt":[{"subitem_source_identifier":"1882-7764","subitem_source_identifier_type":"ISSN"}]},"item_2_description_7":{"attribute_name":"論文抄録","attribute_value_mlt":[{"subitem_description":"本論文では，トランスダクティブ・ブースティング法によるテキスト分類手法を提案する．テキスト分類器の学習に使用する大規模な訓練データの作成にはコストや時間がかかる．そのため訓練データが少ない場合にも高い分類精度が得られる学習法が求められている．トランスダクティブ法は学習の際に訓練データだけでなく，分類クラスの付与されていないテストデータの分布も考慮に入れることにより分類精度を上げる方法である．本論文ではこれをブースティングに対し適用し，実験を行った．その結果，従来のブースティングによる学習に比べて高精度のテキスト分類器を学習できた．特に少数の訓練データしかない場合にも高い精度が得られた．","subitem_description_type":"Other"}]},"item_2_description_8":{"attribute_name":"論文抄録(英)","attribute_value_mlt":[{"subitem_description":"This paper describes a new text categorization method using transductiveboosting.  It is time-consuming and expensive to assemble a large corpus of categorized textfor use with learning-based classification methods.Therefore, we require learning methods that are able to learn classifiersextremely accurately from a small quantity of training data.The transductive method takes account of bothtraining data and test data distribution and provides a highly accurate classifier.We adopt a transductive method in a boosting algorithm for text categorization. The categorization performance was better than  that of the original boosting.Specifically the performance wasimproved significantly for small quantities of training data.","subitem_description_type":"Other"}]},"item_2_biblio_info_10":{"attribute_name":"書誌情報","attribute_value_mlt":[{"bibliographicPageEnd":"1851","bibliographic_titles":[{"bibliographic_title":"情報処理学会論文誌"}],"bibliographicPageStart":"1843","bibliographicIssueDates":{"bibliographicIssueDate":"2002-06-15","bibliographicIssueDateType":"Issued"},"bibliographicIssueNumber":"6","bibliographicVolumeNumber":"43"}]},"relation_version_is_last":true,"item_2_alternative_title_2":{"attribute_name":"その他タイトル","attribute_value_mlt":[{"subitem_alternative_title":"自然言語処理"}]},"weko_creator_id":"1"},"id":11605,"updated":"2025-01-23T02:07:53.405942+00:00","links":{},"created":"2025-01-18T22:46:11.483540+00:00"}