{"metadata":{"_oai":{"id":"oai:ipsj.ixsq.nii.ac.jp:00048458","sets":["1164:4179:4229:4230"]},"path":["4230"],"owner":"1","recid":"48458","title":["Support Vector Machineの多値分類問題への適用法について"],"pubdate":{"attribute_name":"公開日","attribute_value":"2001-11-20"},"_buckets":{"deposit":"98106784-30f1-4ce2-9ca0-7cb9ec1b4e7f"},"_deposit":{"id":"48458","pid":{"type":"depid","value":"48458","revision_id":0},"owners":[1],"status":"published","created_by":1},"item_title":"Support Vector Machineの多値分類問題への適用法について","author_link":["0","0"],"item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"Support Vector Machineの多値分類問題への適用法について"},{"subitem_title":"Applying Support Vector Machine to Multi - Class Classification Problems","subitem_title_language":"en"}]},"item_type_id":"4","publish_date":"2001-11-20","item_4_text_3":{"attribute_name":"著者所属","attribute_value_mlt":[{"subitem_text_value":"奈良先端科学技術大学院大学情報科学研究科"},{"subitem_text_value":"奈良先端科学技術大学院大学情報科学研究科"}]},"item_4_text_4":{"attribute_name":"著者所属(英)","attribute_value_mlt":[{"subitem_text_value":"Graduate School of Information Science, Nara Institute Science and Technology","subitem_text_language":"en"},{"subitem_text_value":"Graduate School of Information Science, Nara Institute Science and Technology","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/48458/files/IPSJ-NL01146006.pdf"},"date":[{"dateType":"Available","dateValue":"2003-11-20"}],"format":"application/pdf","billing":["billing_file"],"filename":"IPSJ-NL01146006.pdf","filesize":[{"value":"177.3 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":"23"},{"tax":["include_tax"],"price":"0","billingrole":"44"}],"accessrole":"open_date","version_id":"162642f6-c0f1-4056-9249-8d2e0b44e733","displaytype":"detail","licensetype":"license_note","license_note":"Copyright (c) 2001 by the Information Processing Society of Japan"}]},"item_4_creator_5":{"attribute_name":"著者名","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"山田, 寛康"},{"creatorName":"松本, 裕治"}],"nameIdentifiers":[{}]}]},"item_4_creator_6":{"attribute_name":"著者名(英)","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"Yamada, Hiroyasu","creatorNameLang":"en"},{"creatorName":"Matsumoto, Yuji","creatorNameLang":"en"}],"nameIdentifiers":[{}]}]},"item_4_source_id_9":{"attribute_name":"書誌レコードID","attribute_value_mlt":[{"subitem_source_identifier":"AN10115061","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":"本研究では  日本語固有表現抽出タスクを題材に  機械学習アルゴリズムSupport Vector Machine(SVM)を多値分類問題に適用する手法を提案し  代表的な従来手法である one vs. rest 法  及び pairwise法 との比較を行なう.  二値分類器であるSVMを固有表現抽出タスクに適用するためには  多値分類器に拡張する必要がある. しかし分類するクラス数に比例して計算コストが増加するため  現実的な時間での学習  及び分類が困難となる. 我々は  多値分類問題を  比較的分類が容易な二値分類へ分割し  二分木を構築する手法を応用し  効率的な学習  及び分類ができるよう  SVMの多値分類器への拡張を行う. 固有表現抽出実験では  従来法である pairwise 法  及び one vs. rest 法と比べ  ほぼ同等な抽出精度を維持し  抽出時間を削減できることを確認した.","subitem_description_type":"Other"}]},"item_4_description_8":{"attribute_name":"論文抄録(英)","attribute_value_mlt":[{"subitem_description":"This paper proposes a method for multi-class classification with Support Vector Machines(SVM) and evaluates its effectiveness using Japanese named entity extraction task. Multi-class problems with more than two classes have typically been solved by combining independently produced binary classifiers, such as pairwise and one vs. rest method. However, these methods require large computational cost with increasing the number of classes. We propose a method to reduce multi-class classification to binary using a method called as tree-structured model for efficient learning and classifying. Results of our extraction experiments suggest that the method is comparable to the one vs. rest and pairwise methods, and it can reduce the extraction time.","subitem_description_type":"Other"}]},"item_4_biblio_info_10":{"attribute_name":"書誌情報","attribute_value_mlt":[{"bibliographicPageEnd":"38","bibliographic_titles":[{"bibliographic_title":"情報処理学会研究報告自然言語処理（NL）"}],"bibliographicPageStart":"33","bibliographicIssueDates":{"bibliographicIssueDate":"2001-11-20","bibliographicIssueDateType":"Issued"},"bibliographicIssueNumber":"112(2001-NL-146)","bibliographicVolumeNumber":"2001"}]},"relation_version_is_last":true,"weko_creator_id":"1"},"updated":"2025-01-22T08:29:33.093445+00:00","created":"2025-01-18T23:13:42.007343+00:00","links":{},"id":48458}