{"id":9266,"created":"2025-01-18T22:44:31.851502+00:00","metadata":{"_oai":{"id":"oai:ipsj.ixsq.nii.ac.jp:00009266","sets":["581:582:584"]},"path":["584"],"owner":"1","recid":"9266","title":["スライド情報を用いた言語モデル適応による講義音声認識"],"pubdate":{"attribute_name":"公開日","attribute_value":"2009-02-15"},"_buckets":{"deposit":"202a7e40-dd36-4afe-9905-22e7aeb97377"},"_deposit":{"id":"9266","pid":{"type":"depid","value":"9266","revision_id":0},"owners":[1],"status":"published","created_by":1},"item_title":"スライド情報を用いた言語モデル適応による講義音声認識","author_link":["0","0"],"item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"スライド情報を用いた言語モデル適応による講義音声認識"},{"subitem_title":"Automatic Lecture Transcription by Exploiting Slide Information for Language Model Adaptation","subitem_title_language":"en"}]},"item_keyword":{"attribute_name":"キーワード","attribute_value_mlt":[{"subitem_subject":"特集：音声ドキュメント処理","subitem_subject_scheme":"Other"}]},"item_type_id":"2","publish_date":"2009-02-15","item_2_text_3":{"attribute_name":"著者所属","attribute_value_mlt":[{"subitem_text_value":"京都大学大学院情報学研究科"},{"subitem_text_value":"京都大学大学院情報学研究科"},{"subitem_text_value":"京都大学大学院情報学研究科"},{"subitem_text_value":"京都大学大学院情報学研究科"}]},"item_2_text_4":{"attribute_name":"著者所属(英)","attribute_value_mlt":[{"subitem_text_value":"Graduate School of Informatics, Kyoto University","subitem_text_language":"en"},{"subitem_text_value":"Graduate School of Informatics, Kyoto University","subitem_text_language":"en"},{"subitem_text_value":"Graduate School of Informatics, Kyoto University","subitem_text_language":"en"},{"subitem_text_value":"Graduate School of Informatics, Kyoto University","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/9266/files/IPSJ-JNL5002004.pdf"},"date":[{"dateType":"Available","dateValue":"2011-02-15"}],"format":"application/pdf","billing":["billing_file"],"filename":"IPSJ-JNL5002004.pdf","filesize":[{"value":"322.5 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":"37015a8c-19cd-436d-8cd7-d654e5967e05","displaytype":"detail","licensetype":"license_note","license_note":"Copyright (c) 2009 by the Information Processing Society of Japan"}]},"item_2_creator_5":{"attribute_name":"著者名","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"河原, 達也"},{"creatorName":"根本, 雄介"},{"creatorName":"勝丸徳浩"},{"creatorName":"秋田, 祐哉"}],"nameIdentifiers":[{}]}]},"item_2_creator_6":{"attribute_name":"著者名(英)","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"Tatsuya, Kawahara","creatorNameLang":"en"},{"creatorName":"Yusuke, Nemoto","creatorNameLang":"en"},{"creatorName":"Norihiro, Katsumaru","creatorNameLang":"en"},{"creatorName":"Yuya, Akita","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":"大学などの講義で使用されるスライドの情報を用いて，言語モデルを動的に適応することにより，音声認識の高精度化を実現する方法を提案する．まず，当該講義のスライド全体のテキストを用いて，PLSA（Probabilistic Latent Semantic Analysis）によりN-gramモデルの話題への適応を行う．次に，発話に対応する個々のスライドの情報を用いて，キャッシュモデルによりスライドに現れる単語の確率を強化し，認識結果のリスコアリングを行う．京都大学で行われた技術講習会と正規の講義を対象とした音声認識において評価を行った結果，PLSAによる大域的な適応とキャッシュモデルによる局所的な適応を組み合わせることにより，認識精度の有意な改善が得られた．特に，キーワードの検出で大きな改善が得られ，大学の講義でも80%に近い精度（F値）を実現した．","subitem_description_type":"Other"}]},"item_2_description_8":{"attribute_name":"論文抄録(英)","attribute_value_mlt":[{"subitem_description":"We investigate several language model adaptation methods which exploit presentation slide information for automatic lecture transcription. First, N-gram probabilities are re-scaled with lecture-dependent unigram probabilities estimated by PLSA (Probabilistic Latent Semantic Analysis) using all slides of the lecture. Then, N-best hypotheses of the initial speech recognition results are re-scored using word probabilities enhanced with a cache model using the slide corresponding to each utterance. Experimental evaluations on real lectures show that the proposed method with the combination of the global and local slide information achieves a significant improvement of recognition accuracy, especially in the detection rate of content keywords.","subitem_description_type":"Other"}]},"item_2_biblio_info_10":{"attribute_name":"書誌情報","attribute_value_mlt":[{"bibliographicPageEnd":"476","bibliographic_titles":[{"bibliographic_title":"情報処理学会論文誌"}],"bibliographicPageStart":"469","bibliographicIssueDates":{"bibliographicIssueDate":"2009-02-15","bibliographicIssueDateType":"Issued"},"bibliographicIssueNumber":"2","bibliographicVolumeNumber":"50"}]},"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":"言語モデル、トピック推定"}]},"weko_creator_id":"1"},"updated":"2025-01-23T03:27:58.610814+00:00","links":{}}