{"id":209902,"updated":"2025-01-19T18:21:06.152430+00:00","links":{},"created":"2025-01-19T01:11:11.059270+00:00","metadata":{"_oai":{"id":"oai:ipsj.ixsq.nii.ac.jp:00209902","sets":["1164:8666:10537:10538"]},"path":["10538"],"owner":"44499","recid":"209902","title":["書誌データ・青空文庫・点字データを用いた振り仮名注釈付き日本語コーパスの構築"],"pubdate":{"attribute_name":"公開日","attribute_value":"2021-02-26"},"_buckets":{"deposit":"1520c98e-6664-478d-8916-bb51619e9a63"},"_deposit":{"id":"209902","pid":{"type":"depid","value":"209902","revision_id":0},"owners":[44499],"status":"published","created_by":44499},"item_title":"書誌データ・青空文庫・点字データを用いた振り仮名注釈付き日本語コーパスの構築","author_link":["530355","530354","530353","530358","530356","530357"],"item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"書誌データ・青空文庫・点字データを用いた振り仮名注釈付き日本語コーパスの構築"},{"subitem_title":"Construction of a Japanese corpus with furigana annotations using bibliographic data, Aozora Bunko, and Braille data","subitem_title_language":"en"}]},"item_keyword":{"attribute_name":"キーワード","attribute_value_mlt":[{"subitem_subject":"コンテンツのアクセシビリティ","subitem_subject_scheme":"Other"}]},"item_type_id":"4","publish_date":"2021-02-26","item_4_text_3":{"attribute_name":"著者所属","attribute_value_mlt":[{"subitem_text_value":"東京大学大学院情報理工学系研究科"},{"subitem_text_value":"東京大学生産技術研究所"},{"subitem_text_value":"国立情報学研究所／東京大学生産技術研究所"}]},"item_4_text_4":{"attribute_name":"著者所属(英)","attribute_value_mlt":[{"subitem_text_value":"Graduate School of Information Science and Technology, the University of Tokyo","subitem_text_language":"en"},{"subitem_text_value":"Institute of Industrial Science, the University of Tokyo","subitem_text_language":"en"},{"subitem_text_value":"National Institute of Informatics / Institute of Industrial Science, the University of Tokyo","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/209902/files/IPSJ-AAC21015013.pdf","label":"IPSJ-AAC21015013.pdf"},"date":[{"dateType":"Available","dateValue":"2023-02-26"}],"format":"application/pdf","billing":["billing_file"],"filename":"IPSJ-AAC21015013.pdf","filesize":[{"value":"1.1 MB"}],"mimetype":"application/pdf","priceinfo":[{"tax":["include_tax"],"price":"660","billingrole":"5"},{"tax":["include_tax"],"price":"330","billingrole":"6"},{"tax":["include_tax"],"price":"0","billingrole":"52"},{"tax":["include_tax"],"price":"0","billingrole":"44"}],"accessrole":"open_date","version_id":"c467267a-a078-4246-87f4-a3eb007aa7f9","displaytype":"detail","licensetype":"license_note","license_note":"Copyright (c) 2021 by the Information Processing Society of Japan"}]},"item_4_creator_5":{"attribute_name":"著者名","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"佐藤, 文一"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"吉永, 直樹"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"喜連川, 優"}],"nameIdentifiers":[{}]}]},"item_4_creator_6":{"attribute_name":"著者名(英)","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"Fumikazu, Sato","creatorNameLang":"en"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"Naoki, Yoshinaga","creatorNameLang":"en"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"Masaru, Kitsuregawa","creatorNameLang":"en"}],"nameIdentifiers":[{}]}]},"item_4_source_id_9":{"attribute_name":"書誌レコードID","attribute_value_mlt":[{"subitem_source_identifier":"AA12752949","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":"2432-2431","subitem_source_identifier_type":"ISSN"}]},"item_4_description_7":{"attribute_name":"論文抄録","attribute_value_mlt":[{"subitem_description":"重度視覚障害者は，パソコンで画面読み上げソフトを使用して，漢字交じりの文書を音声で聞いているが，しばしば読み誤りが発生する．例えば，「表に出る」を「ひょうにでる」と読み上げられると理解が困難になる．この問題に対しては，機械学習に基づく統計的手法を用いて，前述の表（ひょう，おもて）のような同形異音語の読みを推定するアプローチが有望であるが，モデルの学習には正しい振り仮名が付いた大量の文が必要になる．そこで我々は，振り仮名が付与された国立国会図書館の書誌データの雑誌タイトルや，校正済みの点字データなどを活用して，機械学習に基づく読み推定モデルの学習に必要となる振り仮名付きの日本語コーパスを構築した．具体的には，まず，書誌データのタイトルとその振り仮名のペア，青空文庫のテキストと官公庁が公開する障害者向けの広報テキストなどの PDF テキストと，それらに該当する点字のデータから，対応する文のペアをパターンマッチングで選び出す．次に，この文対に対して，既存の振り仮名注釈付きコーパスや形態素解析辞書などから事前に収集した漢字に対する振り仮名候補に基づく文字レベルのマッチングを行い，振り仮名注釈付き日本語コーパスを構築する．約 4.1 億文字の漢字仮名交じり文の中から，約 3.5 億文字の文に含まれるすべての漢字に対して振り仮名の自動注釈を行った．これにより同形異音語を含む文を選び出せることを確認した．","subitem_description_type":"Other"}]},"item_4_biblio_info_10":{"attribute_name":"書誌情報","attribute_value_mlt":[{"bibliographicPageEnd":"8","bibliographic_titles":[{"bibliographic_title":"研究報告アクセシビリティ（AAC）"}],"bibliographicPageStart":"1","bibliographicIssueDates":{"bibliographicIssueDate":"2021-02-26","bibliographicIssueDateType":"Issued"},"bibliographicIssueNumber":"13","bibliographicVolumeNumber":"2021-AAC-15"}]},"relation_version_is_last":true,"weko_creator_id":"44499"}}