{"updated":"2025-01-19T08:51:05.945675+00:00","links":{},"metadata":{"_oai":{"id":"oai:ipsj.ixsq.nii.ac.jp:00237542","sets":["934:1022:11484:11667"]},"path":["11667"],"owner":"44499","recid":"237542","title":["Acceptability Evaluation of Naturally Written Sentences"],"pubdate":{"attribute_name":"公開日","attribute_value":"2024-07-24"},"_buckets":{"deposit":"b5c5e880-b4bd-4dbd-9a86-572db44c6c09"},"_deposit":{"id":"237542","pid":{"type":"depid","value":"237542","revision_id":0},"owners":[44499],"status":"published","created_by":44499},"item_title":"Acceptability Evaluation of Naturally Written Sentences","author_link":["650427","650429","650430","650431","650428","650426"],"item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"Acceptability Evaluation of Naturally Written Sentences"},{"subitem_title":"Acceptability Evaluation of Naturally Written Sentences","subitem_title_language":"en"}]},"item_keyword":{"attribute_name":"キーワード","attribute_value_mlt":[{"subitem_subject":"[研究論文] acceptability, readability, grammaticality, generative text, text evaluation, syntactic knowledge, speakers judgement","subitem_subject_scheme":"Other"}]},"item_type_id":"3","publish_date":"2024-07-24","item_3_text_3":{"attribute_name":"著者所属","attribute_value_mlt":[{"subitem_text_value":"Tokyo Institute of Technology"},{"subitem_text_value":"University of Pennsylvania"},{"subitem_text_value":"Tokyo Institute of Technology"}]},"item_3_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":"University of Pennsylvania","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/237542/files/IPSJ-TOD1703002.pdf","label":"IPSJ-TOD1703002.pdf"},"date":[{"dateType":"Available","dateValue":"2026-07-24"}],"format":"application/pdf","billing":["billing_file"],"filename":"IPSJ-TOD1703002.pdf","filesize":[{"value":"901.2 kB"}],"mimetype":"application/pdf","priceinfo":[{"tax":["include_tax"],"price":"0","billingrole":"5"},{"tax":["include_tax"],"price":"0","billingrole":"6"},{"tax":["include_tax"],"price":"0","billingrole":"13"},{"tax":["include_tax"],"price":"0","billingrole":"39"},{"tax":["include_tax"],"price":"0","billingrole":"44"}],"accessrole":"open_date","version_id":"7522e2f8-0928-4152-8e5f-73a52be64cfc","displaytype":"detail","licensetype":"license_note","license_note":"Copyright (c) 2024 by the Information Processing Society of Japan"}]},"item_3_creator_5":{"attribute_name":"著者名","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"Vijay, Daultani"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"Héctor, Javier Vázquez Martínez"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"Naoaki, Okazaki"}],"nameIdentifiers":[{}]}]},"item_3_creator_6":{"attribute_name":"著者名(英)","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"Vijay, Daultani","creatorNameLang":"en"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"Héctor, Javier Vázquez Martínez","creatorNameLang":"en"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"Naoaki, Okazaki","creatorNameLang":"en"}],"nameIdentifiers":[{}]}]},"item_3_source_id_9":{"attribute_name":"書誌レコードID","attribute_value_mlt":[{"subitem_source_identifier":"AA11464847","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_3_source_id_11":{"attribute_name":"ISSN","attribute_value_mlt":[{"subitem_source_identifier":"1882-7799","subitem_source_identifier_type":"ISSN"}]},"item_3_description_7":{"attribute_name":"論文抄録","attribute_value_mlt":[{"subitem_description":"The success of Language Models (LMs) on a variety of NLP tasks has prompted the design and analysis of natural language benchmarks to evaluate their fitness for particular applications. In this work, we focus on the NLP task of acceptability rating, whereby a given model must rate the ‘goodness’ of a series of tokens. We find the current commonly used datasets to benchmark for LM sentence acceptability fail to capture the distribution of naturally occurring written data. Moreover, we find that the bias toward shorter (5-8 word) sentences is a strong confounding factor that contributes positively to LMs' performance. We then introduce seven datasets collected from the NLP literature that closely follow the sentence length distribution of naturally occurring written text. In our experiments, when sentence length is controlled by adjusting the distribution to match naturally occurring data, we observe a performance drop for current commonly used datasets of up to 48 points in MCC. We conclude with a discussion on implications for current applications and recommendations to improve our current commonly used acceptability benchmarking datasets.\n------------------------------\nThis is a preprint of an article intended for publication Journal of\nInformation Processing(JIP). This preprint should not be cited. This\narticle should be cited as: Journal of Information Processing Vol.32(2024) (online)\n------------------------------","subitem_description_type":"Other"}]},"item_3_description_8":{"attribute_name":"論文抄録(英)","attribute_value_mlt":[{"subitem_description":"The success of Language Models (LMs) on a variety of NLP tasks has prompted the design and analysis of natural language benchmarks to evaluate their fitness for particular applications. In this work, we focus on the NLP task of acceptability rating, whereby a given model must rate the ‘goodness’ of a series of tokens. We find the current commonly used datasets to benchmark for LM sentence acceptability fail to capture the distribution of naturally occurring written data. Moreover, we find that the bias toward shorter (5-8 word) sentences is a strong confounding factor that contributes positively to LMs' performance. We then introduce seven datasets collected from the NLP literature that closely follow the sentence length distribution of naturally occurring written text. In our experiments, when sentence length is controlled by adjusting the distribution to match naturally occurring data, we observe a performance drop for current commonly used datasets of up to 48 points in MCC. We conclude with a discussion on implications for current applications and recommendations to improve our current commonly used acceptability benchmarking datasets.\n------------------------------\nThis is a preprint of an article intended for publication Journal of\nInformation Processing(JIP). This preprint should not be cited. This\narticle should be cited as: Journal of Information Processing Vol.32(2024) (online)\n------------------------------","subitem_description_type":"Other"}]},"item_3_biblio_info_10":{"attribute_name":"書誌情報","attribute_value_mlt":[{"bibliographic_titles":[{"bibliographic_title":"情報処理学会論文誌データベース（TOD）"}],"bibliographicIssueDates":{"bibliographicIssueDate":"2024-07-24","bibliographicIssueDateType":"Issued"},"bibliographicIssueNumber":"3","bibliographicVolumeNumber":"17"}]},"relation_version_is_last":true,"weko_creator_id":"44499"},"id":237542,"created":"2025-01-19T01:40:13.627043+00:00"}