{"metadata":{"_oai":{"id":"oai:ipsj.ixsq.nii.ac.jp:00231498","sets":["1164:3500:11130:11413"]},"path":["11413"],"owner":"44499","recid":"231498","title":["BERTopicによる複数の分類器を用いた文書分類の安定性調査"],"pubdate":{"attribute_name":"公開日","attribute_value":"2023-12-13"},"_buckets":{"deposit":"ee4f7bee-d0cb-44fb-bab8-22974c8c509e"},"_deposit":{"id":"231498","pid":{"type":"depid","value":"231498","revision_id":0},"owners":[44499],"status":"published","created_by":44499},"item_title":"BERTopicによる複数の分類器を用いた文書分類の安定性調査","author_link":["625066","625067","625064","625065"],"item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"BERTopicによる複数の分類器を用いた文書分類の安定性調査"},{"subitem_title":"Stability study of document classification with multiple classifiers using BERTopic","subitem_title_language":"en"}]},"item_keyword":{"attribute_name":"キーワード","attribute_value_mlt":[{"subitem_subject":"マスク言語モデル","subitem_subject_scheme":"Other"}]},"item_type_id":"4","publish_date":"2023-12-13","item_4_text_3":{"attribute_name":"著者所属","attribute_value_mlt":[{"subitem_text_value":"現在，工学院大学"},{"subitem_text_value":"現在，工学院大学"}]},"item_4_text_4":{"attribute_name":"著者所属(英)","attribute_value_mlt":[{"subitem_text_value":"Presently with Kogakuin University","subitem_text_language":"en"},{"subitem_text_value":"Presently with Kogakuin University","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/231498/files/IPSJ-IFAT23153008.pdf","label":"IPSJ-IFAT23153008.pdf"},"date":[{"dateType":"Available","dateValue":"2025-12-13"}],"format":"application/pdf","billing":["billing_file"],"filename":"IPSJ-IFAT23153008.pdf","filesize":[{"value":"939.6 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":"39"},{"tax":["include_tax"],"price":"0","billingrole":"44"}],"accessrole":"open_date","version_id":"c2b5bccd-c6d2-4821-9da6-30e2735320c6","displaytype":"detail","licensetype":"license_note","license_note":"Copyright (c) 2023 by the Information Processing Society of Japan"}]},"item_4_creator_5":{"attribute_name":"著者名","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"櫻井, 勇気"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"小林, 亜樹"}],"nameIdentifiers":[{}]}]},"item_4_creator_6":{"attribute_name":"著者名(英)","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"Yuki, Sakurai","creatorNameLang":"en"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"Aki, Kobayashi","creatorNameLang":"en"}],"nameIdentifiers":[{}]}]},"item_4_source_id_9":{"attribute_name":"書誌レコードID","attribute_value_mlt":[{"subitem_source_identifier":"AN10114171","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_source_id_11":{"attribute_name":"ISSN","attribute_value_mlt":[{"subitem_source_identifier":"2188-8884","subitem_source_identifier_type":"ISSN"}]},"item_4_description_7":{"attribute_name":"論文抄録","attribute_value_mlt":[{"subitem_description":"BERTopic は埋め込み表現，次元圧縮，分類，可視化手法の変更を可能とするライブラリを用いたトピックモデリング手法であり，事前に学習された Transformer ベースの言語モデルを用いて文書の埋め込み表現を獲得し，HDBSCAN によるクラスタリングを行う．文書分類タスクにおいて，BERTopic は LDA などの従来の手法よりも高い精度で文書分類を行うことができるが，分類を行うたびにクラスターが変化するため，同一の分類結果を出力することはない．本論文では，BERTopic を用いて異なる埋め込み表現を持つ複数の分類器を作成し，これらを用いて同一データセットに対し分類を行うことで BERTopic が文書集合に対してどれほど安定した結果をもたらすのか調査を行い，結果を考察する．","subitem_description_type":"Other"}]},"item_4_description_8":{"attribute_name":"論文抄録(英)","attribute_value_mlt":[{"subitem_description":"BERTopic is a topic modeling method using a library that allows modification of embedded representations, dimensionality compression, classification, and visualization methods. It acquires embedded representations of documents using pre-trained Transformer-based language models and performs clustering using HDBSCAN. In the document classification task, BERTopic can classify documents with higher accuracy than conventional methods such as LDA. Still, it does not output identical classification results because the clusters change with each classification. In this paper, we investigate how BERTopic provides stable results for a set of documents by creating multiple classifiers with different embedding representations using BERTopic and performing classification on the same dataset using these classifiers.","subitem_description_type":"Other"}]},"item_4_biblio_info_10":{"attribute_name":"書誌情報","attribute_value_mlt":[{"bibliographicPageEnd":"4","bibliographic_titles":[{"bibliographic_title":"研究報告情報基礎とアクセス技術（IFAT）"}],"bibliographicPageStart":"1","bibliographicIssueDates":{"bibliographicIssueDate":"2023-12-13","bibliographicIssueDateType":"Issued"},"bibliographicIssueNumber":"8","bibliographicVolumeNumber":"2023-IFAT-153"}]},"relation_version_is_last":true,"weko_creator_id":"44499"},"id":231498,"updated":"2025-01-19T10:44:25.688341+00:00","links":{},"created":"2025-01-19T01:31:50.271778+00:00"}