{"updated":"2025-01-19T14:38:19.877045+00:00","links":{},"id":220233,"created":"2025-01-19T01:20:16.731451+00:00","metadata":{"_oai":{"id":"oai:ipsj.ixsq.nii.ac.jp:00220233","sets":["1164:4179:10952:11016"]},"path":["11016"],"owner":"44499","recid":"220233","title":["疑似訓練データを用いたBERTによる同形異音語の読み推定"],"pubdate":{"attribute_name":"公開日","attribute_value":"2022-09-22"},"_buckets":{"deposit":"36a82953-4168-4de2-a142-71ab4501f0e9"},"_deposit":{"id":"220233","pid":{"type":"depid","value":"220233","revision_id":0},"owners":[44499],"status":"published","created_by":44499},"item_title":"疑似訓練データを用いたBERTによる同形異音語の読み推定","author_link":["575573","575574","575572"],"item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"疑似訓練データを用いたBERTによる同形異音語の読み推定"}]},"item_keyword":{"attribute_name":"キーワード","attribute_value_mlt":[{"subitem_subject":"解析","subitem_subject_scheme":"Other"}]},"item_type_id":"4","publish_date":"2022-09-22","item_4_text_3":{"attribute_name":"著者所属","attribute_value_mlt":[{"subitem_text_value":"茨城大学大学院理工学研究科情報工学専攻"},{"subitem_text_value":"東京農工大学大学院工学研究院先端情報科学部門"},{"subitem_text_value":"茨城大学大学院理工学研究科情報科学領域"}]},"item_4_text_4":{"attribute_name":"著者所属(英)","attribute_value_mlt":[{"subitem_text_value":"Major in Computer and Information Sciences, Graduate School of Science and Engineering, Ibaraki University","subitem_text_language":"en"},{"subitem_text_value":"Division of Advanced Information Technology & Computer Science, Institute of Engineering, Tokyo University of Agriculture and Technology","subitem_text_language":"en"},{"subitem_text_value":"Graduate School of Science and Engineering, Department of Computer and Information Sciences, Ibaraki 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/220233/files/IPSJ-NL22253003.pdf","label":"IPSJ-NL22253003.pdf"},"date":[{"dateType":"Available","dateValue":"2024-09-22"}],"format":"application/pdf","billing":["billing_file"],"filename":"IPSJ-NL22253003.pdf","filesize":[{"value":"935.3 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":"adb4d782-5af1-49eb-aeb0-3b5383dcd9c5","displaytype":"detail","licensetype":"license_note","license_note":"Copyright (c) 2022 by the Information Processing Society of Japan"}]},"item_4_creator_5":{"attribute_name":"著者名","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"小林, 汰一郎"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"古宮, 嘉那子"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"新納, 浩幸"}],"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_source_id_11":{"attribute_name":"ISSN","attribute_value_mlt":[{"subitem_source_identifier":"2188-8779","subitem_source_identifier_type":"ISSN"}]},"item_4_description_7":{"attribute_name":"論文抄録","attribute_value_mlt":[{"subitem_description":"日本語には読みに曖昧性を持つ単語が多数存在する．例えば「辛い」は「カライ」のほかに「ツライ」と読むこともできる．このような単語を同形異音語と呼ぶ．本論文では，BERT を用いて同形異音語の読み推定を行う．訓練・テストデータには現代日本語書き言葉均衡コーパス (BCCWJ) と日本語話し言葉コーパス (CSJ) を利用した．BCCWJ の大半を占める非コアデータの読みは，形態素解析システム MeCab により機械的に割り振られたものである．また，BCCWJ は書き言葉であり，CSJ は話し言葉なので，ドメインのずれが想定される．CSJ をターゲット領域としたとき，通常はこの領域の訓練事例を用いて読み推定のモデルを学習・構築すればよいが，訓練事例の構築コストが高いという問題がある．本研究では自動的に付与されたドメイン外の大量の疑似データ (BCCWJ のデータ) を利用することで，本来必要としたターゲットの領域の訓練事例の量を大幅に削減することができた．","subitem_description_type":"Other"}]},"item_4_biblio_info_10":{"attribute_name":"書誌情報","attribute_value_mlt":[{"bibliographicPageEnd":"5","bibliographic_titles":[{"bibliographic_title":"研究報告自然言語処理（NL）"}],"bibliographicPageStart":"1","bibliographicIssueDates":{"bibliographicIssueDate":"2022-09-22","bibliographicIssueDateType":"Issued"},"bibliographicIssueNumber":"3","bibliographicVolumeNumber":"2022-NL-253"}]},"relation_version_is_last":true,"weko_creator_id":"44499"}}