{"links":{},"id":57095,"metadata":{"_oai":{"id":"oai:ipsj.ixsq.nii.ac.jp:00057095","sets":["1164:5159:5186:5188"]},"path":["5188"],"owner":"1","recid":"57095","title":["『日本語話し言葉コーパス』を用いた汎用的な発音変動モデルの統計的学習"],"pubdate":{"attribute_name":"公開日","attribute_value":"2004-10-22"},"_buckets":{"deposit":"5d445977-c83e-4771-bb0a-0ede74078e9c"},"_deposit":{"id":"57095","pid":{"type":"depid","value":"57095","revision_id":0},"owners":[1],"status":"published","created_by":1},"item_title":"『日本語話し言葉コーパス』を用いた汎用的な発音変動モデルの統計的学習","author_link":["0","0"],"item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"『日本語話し言葉コーパス』を用いた汎用的な発音変動モデルの統計的学習"},{"subitem_title":"Generalized Statistical Modeling of Pronunciation Variations using the Corpus of Spontaneous Japanese","subitem_title_language":"en"}]},"item_type_id":"4","publish_date":"2004-10-22","item_4_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／科学技術振興機構さきがけ研究21"},{"subitem_text_value":"京都大学大学院情報学研究科／科学技術振興機構さきがけ研究21"}]},"item_4_text_4":{"attribute_name":"著者所属(英)","attribute_value_mlt":[{"subitem_text_value":"School of Informatics, Kyoto University/PRESTO, Japan Science and Technology Agency (JST)","subitem_text_language":"en"},{"subitem_text_value":"School of Informatics, Kyoto University/PRESTO, Japan Science and Technology Agency (JST)","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/57095/files/IPSJ-SLP04053003.pdf","label":"IPSJ-SLP04053003"},"date":[{"dateType":"Available","dateValue":"2006-10-22"}],"format":"application/pdf","billing":["billing_file"],"filename":"IPSJ-SLP04053003.pdf","filesize":[{"value":"175.9 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":"22"},{"tax":["include_tax"],"price":"0","billingrole":"44"}],"accessrole":"open_date","version_id":"c0d69f2b-8d7e-4e0d-b375-f0185584e7b0","displaytype":"detail","licensetype":"license_note","license_note":"Copyright (c) 2004 by the Information Processing Society of Japan"}]},"item_4_creator_5":{"attribute_name":"著者名","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"秋田, 祐哉"},{"creatorName":"河原, 達也"}],"nameIdentifiers":[{}]}]},"item_4_creator_6":{"attribute_name":"著者名(英)","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"Yuya, Akita","creatorNameLang":"en"},{"creatorName":"Tatsuya, Kawahara","creatorNameLang":"en"}],"nameIdentifiers":[{}]}]},"item_4_source_id_9":{"attribute_name":"書誌レコードID","attribute_value_mlt":[{"subitem_source_identifier":"AN10442647","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":"話し言葉音声の認識において，発音変動のモデル化は認識性能に深く関わる課題である．通常，音声認識に用いる発音辞書は形態素解析器が出力する標準的な読みに基づいて生成されるが，これでは話し言葉に多く含まれる発音変動をカバーできない．本研究では，まず『日本語話し言葉コーパス』（CSJ）を用いて発音変動のパターンを汎用的な音素系列のレベルで統計的に学習した．コーパスから自動的に獲得された音素列の変動パターンは265種類であり，音韻論的に妥当なものに加えて人手による規則化が困難なものを頻度統計とあわせて抽出することができた．これらのパターンに対して，バックオフ手法により可変長の音素文脈を扱える確率つき音素書き換え規則を構築する．これらの規則を適用することで，任意の語いに対して標準的な読み（baseform）から話し言葉特有の変動を含んだ発音（surface form）を生起確率とともに生成することができる．本手法をCSJとは異なるドメインのための発音辞書に適用したところ，エントリ数が21%増加した．さらに，この発音辞書を用いた音声認識により有意な単語誤り率の改善を得ることができた．","subitem_description_type":"Other"}]},"item_4_description_8":{"attribute_name":"論文抄録(英)","attribute_value_mlt":[{"subitem_description":"Pronunciation variation modeling is one of major issues in automatic transcription of spontaneous speech.  We present statistical modeling of subword-based mapping between baseforms and surface forms using a large-scale spontaneous speech corpus (CSJ).  Variation patterns of phone sequences are automatically extracted together with their contexts of up to two preceding and following phones, which are decided by their occurrence statistics.  Then, we derive a set of rewrite rules with their probabilities and variable-length phone contexts.  The model effectively predicts pronunciation variations depending on the phone context using a back-off scheme.  Since it is based on phone sequences, the model is applicable to any lexicon togenerate appropriate surface forms.  The proposed method was evaluatedon a transcription task whose domain is different from the training corpus (CSJ), and significant reduction of word error rate was achieved.","subitem_description_type":"Other"}]},"item_4_biblio_info_10":{"attribute_name":"書誌情報","attribute_value_mlt":[{"bibliographicPageEnd":"18","bibliographic_titles":[{"bibliographic_title":"情報処理学会研究報告音声言語情報処理（SLP）"}],"bibliographicPageStart":"13","bibliographicIssueDates":{"bibliographicIssueDate":"2004-10-22","bibliographicIssueDateType":"Issued"},"bibliographicIssueNumber":"103(2004-SLP-053)","bibliographicVolumeNumber":"2004"}]},"relation_version_is_last":true,"weko_creator_id":"1"},"created":"2025-01-18T23:20:22.379601+00:00","updated":"2025-01-21T15:29:43.893022+00:00"}