{"created":"2025-01-19T01:17:08.126288+00:00","updated":"2025-01-19T15:47:40.264660+00:00","metadata":{"_oai":{"id":"oai:ipsj.ixsq.nii.ac.jp:00216611","sets":["1164:5159:10869:10870"]},"path":["10870"],"owner":"44499","recid":"216611","title":["Incorporating Acoustic and Textual Information for Language Modeling in Code-switching Speech Recognition"],"pubdate":{"attribute_name":"公開日","attribute_value":"2022-02-22"},"_buckets":{"deposit":"8f934eba-70c9-467e-bca1-bf87d2e82aaf"},"_deposit":{"id":"216611","pid":{"type":"depid","value":"216611","revision_id":0},"owners":[44499],"status":"published","created_by":44499},"item_title":"Incorporating Acoustic and Textual Information for Language Modeling in Code-switching Speech Recognition","author_link":["559236","559237","559235","559239","559238","559240"],"item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"Incorporating Acoustic and Textual Information for Language Modeling in Code-switching Speech Recognition"},{"subitem_title":"Incorporating Acoustic and Textual Information for Language Modeling in Code-switching Speech Recognition","subitem_title_language":"en"}]},"item_keyword":{"attribute_name":"キーワード","attribute_value_mlt":[{"subitem_subject":"SP1","subitem_subject_scheme":"Other"}]},"item_type_id":"4","publish_date":"2022-02-22","item_4_text_3":{"attribute_name":"著者所属","attribute_value_mlt":[{"subitem_text_value":"Tokyo Institute of Technology"},{"subitem_text_value":"Tokyo Institute of Technology"},{"subitem_text_value":"Tokyo Institute of Technology"}]},"item_4_text_4":{"attribute_name":"著者所属(英)","attribute_value_mlt":[{"subitem_text_value":"Tokyo Institute of Technology","subitem_text_language":"en"},{"subitem_text_value":"Tokyo Institute of Technology","subitem_text_language":"en"},{"subitem_text_value":"Tokyo Institute of Technology","subitem_text_language":"en"}]},"item_language":{"attribute_name":"言語","attribute_value_mlt":[{"subitem_language":"eng"}]},"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/216611/files/IPSJ-SLP22140010.pdf","label":"IPSJ-SLP22140010.pdf"},"format":"application/pdf","billing":["billing_file"],"filename":"IPSJ-SLP22140010.pdf","filesize":[{"value":"1.8 MB"}],"mimetype":"application/pdf","priceinfo":[{"tax":["include_tax"],"price":"0","billingrole":"22"},{"tax":["include_tax"],"price":"0","billingrole":"44"}],"accessrole":"open_login","version_id":"877b7e85-76cd-4f20-900c-d5a025196115","displaytype":"detail","licensetype":"license_note","license_note":"Copyright (c) 2022 by the Institute of Electronics, Information and Communication Engineers This SIG report is only available to those in membership of the SIG."}]},"item_4_creator_5":{"attribute_name":"著者名","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"Roland, Hartanto"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"Kuniaki, Uto"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"Koichi, Shinoda"}],"nameIdentifiers":[{}]}]},"item_4_creator_6":{"attribute_name":"著者名(英)","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"Roland, Hartanto","creatorNameLang":"en"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"Kuniaki, Uto","creatorNameLang":"en"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"Koichi, Shinoda","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_source_id_11":{"attribute_name":"ISSN","attribute_value_mlt":[{"subitem_source_identifier":"2188-8663","subitem_source_identifier_type":"ISSN"}]},"item_4_description_7":{"attribute_name":"論文抄録","attribute_value_mlt":[{"subitem_description":"People who speak two or more languages tend to alternate the language when they are speaking. This particular phenomenon is called code-switching, and it frequently occurs in multicultural society. Automatic speech recognition (ASR) for code-switching speech is a challenging task acoustically and linguistically because of the lack of code-switching data. This work aims to improve code-switching ASR system by improving the language model. We explore the code-switching data augmentation for language modeling by utilizing the ASR decoding lattice to tackle the pronunciation variation and data scarcity problems. We incorporate both acoustic and textual information by pretraining GPT2, a transformer-based language model, with the code-switching ASR decoding lattice. Our work achieves around 2 point absolute word error rate reduction from the baseline n-gram language model, and 0.33 point absolute reduction from the lattice-rescored baseline word error rate.","subitem_description_type":"Other"}]},"item_4_description_8":{"attribute_name":"論文抄録(英)","attribute_value_mlt":[{"subitem_description":"People who speak two or more languages tend to alternate the language when they are speaking. This particular phenomenon is called code-switching, and it frequently occurs in multicultural society. Automatic speech recognition (ASR) for code-switching speech is a challenging task acoustically and linguistically because of the lack of code-switching data. This work aims to improve code-switching ASR system by improving the language model. We explore the code-switching data augmentation for language modeling by utilizing the ASR decoding lattice to tackle the pronunciation variation and data scarcity problems. We incorporate both acoustic and textual information by pretraining GPT2, a transformer-based language model, with the code-switching ASR decoding lattice. Our work achieves around 2 point absolute word error rate reduction from the baseline n-gram language model, and 0.33 point absolute reduction from the lattice-rescored baseline word error rate.","subitem_description_type":"Other"}]},"item_4_biblio_info_10":{"attribute_name":"書誌情報","attribute_value_mlt":[{"bibliographicPageEnd":"8","bibliographic_titles":[{"bibliographic_title":"研究報告音声言語情報処理（SLP）"}],"bibliographicPageStart":"1","bibliographicIssueDates":{"bibliographicIssueDate":"2022-02-22","bibliographicIssueDateType":"Issued"},"bibliographicIssueNumber":"10","bibliographicVolumeNumber":"2022-SLP-140"}]},"relation_version_is_last":true,"weko_creator_id":"44499"},"id":216611,"links":{}}