Several studies on automated text simplification are based on a large-scale monolingual parallel corpus constructed from a comparable corpus comprising complex text and simple text. However, constructing a parallel corpus for text simplification is expensive as large-scale simplified corpora are not available in many languages other than English. Therefore, we propose an unsupervised method that automatically builds a pseudo-parallel corpus to train a text simplification model. Our framework combines readability assessment and sentence alignment and automatically constructs a text simplification corpus from only a raw corpus. Experimental results show that a statistical machine translation model trained using our corpus can generate simpler synonymous sentences performing comparably to models trained using a large-scale simplified corpus.
"},{"@language":"ja","@value":"難解なテキストと平易なテキストからなる大規模な単言語パラレルコーパスを用いて，テキスト平易化が活発に研究されている．しかし，英語以外の多くの言語では平易に書かれた大規模なコーパスを利用できないため，テキスト平易化のためのパラレルコーパスを構築するコストが高い．そこで本研究では，テキスト平易化のための大規模な疑似パラレルコーパスを自動構築する教師なし手法を提案する．我々の提案するフレームワークでは，リーダビリティ推定と文アライメントを組み合わせることによって，生コーパスのみからテキスト平易化のための単言語パラレルコーパスを自動構築する．統計的機械翻訳を用いた実験の結果，生コーパスのみを用いて学習した我々のテキスト平易化モデルは，平易に書かれた大規模なコーパスを用いて学習した従来のテキスト平易化モデルと同等の性能で平易な同義文を生成できた．
"}],"abstractLicenseFlag":"disallow"}],"creator":[{"@id":"https://cir.nii.ac.jp/crid/1420564276165352704","@type":"Researcher","personIdentifier":[{"@type":"KAKEN_RESEARCHERS","@value":"60581329"},{"@type":"NRID","@value":"1000060581329"},{"@type":"NRID","@value":"9000403320512"},{"@type":"NRID","@value":"9000406337026"},{"@type":"NRID","@value":"9000405868335"},{"@type":"NRID","@value":"9000402795275"},{"@type":"NRID","@value":"9000410015273"},{"@type":"NRID","@value":"9000412365361"},{"@type":"NRID","@value":"9000402795352"},{"@type":"NRID","@value":"9000414941544"},{"@type":"NRID","@value":"9000403793636"},{"@type":"NRID","@value":"9000412365355"},{"@type":"NRID","@value":"9000402113465"},{"@type":"NRID","@value":"9000403674564"},{"@type":"NRID","@value":"9000402796047"},{"@type":"NRID","@value":"9000411647076"},{"@type":"NRID","@value":"9000412357509"},{"@type":"NRID","@value":"9000414836624"},{"@type":"NRID","@value":"9000402796011"},{"@type":"NRID","@value":"9000405705160"},{"@type":"NRID","@value":"9000414941469"},{"@type":"NRID","@value":"9000399224123"},{"@type":"RESEARCHMAP","@value":"https://researchmap.jp/komachi"}],"foaf:name":[{"@language":"ja","@value":"小町 守"},{"@language":"en","@value":"Komachi Mamoru"}],"jpcoar:affiliationName":[{"@language":"en","@value":"Tokyo Metropolitan University"},{"@language":"ja","@value":"首都大学東京"}]},{"@id":"https://cir.nii.ac.jp/crid/1420564276188985472","@type":"Researcher","personIdentifier":[{"@type":"KAKEN_RESEARCHERS","@value":"70824960"},{"@type":"NRID","@value":"1000070824960"},{"@type":"NRID","@value":"9000410181555"},{"@type":"NRID","@value":"9000408453427"},{"@type":"NRID","@value":"9000412365377"},{"@type":"NRID","@value":"9000411647143"},{"@type":"NRID","@value":"9000412365360"},{"@type":"NRID","@value":"9000397672566"},{"@type":"NRID","@value":"9000281478867"},{"@type":"NRID","@value":"9000414941545"},{"@type":"NRID","@value":"9000403957209"},{"@type":"NRID","@value":"9000403674663"},{"@type":"NRID","@value":"9000411647073"},{"@type":"NRID","@value":"9000408451515"},{"@type":"NRID","@value":"9000405705159"},{"@type":"NRID","@value":"9000414941470"},{"@type":"NRID","@value":"9000396134711"},{"@type":"RESEARCHMAP","@value":"https://researchmap.jp/moguranosenshi"}],"foaf:name":[{"@language":"ja","@value":"梶原 智之"},{"@language":"en","@value":"Kajiwara Tomoyuki"}],"jpcoar:affiliationName":[{"@language":"en","@value":"Tokyo Metropolitan University"},{"@language":"ja","@value":"首都大学東京"}]}],"publication":{"publicationIdentifier":[{"@type":"PISSN","@value":"13407619"},{"@type":"LISSN","@value":"13407619"},{"@type":"EISSN","@value":"21858314"},{"@type":"NDL_BIB_ID","@value":"000000093843"},{"@type":"ISSN","@value":"13407619"},{"@type":"NCID","@value":"AN10472659"}],"prism:publicationName":[{"@language":"en","@value":"Journal of Natural Language Processing"},{"@language":"ja","@value":"自然言語処理"},{"@language":"en","@value":"Journal of Natural Language Processing"},{"@language":"ja","@value":"自然言語処理"}],"dc:publisher":[{"@language":"en","@value":"The Association for Natural Language Processing"},{"@language":"ja","@value":"一般社団法人 言語処理学会"}],"prism:publicationDate":"2018-03-15","prism:volume":"25","prism:number":"2","prism:startingPage":"223","prism:endingPage":"249"},"reviewed":"false","dcterms:accessRights":"http://purl.org/coar/access_right/c_abf2","url":[{"@id":"http://id.ndl.go.jp/bib/028919009"},{"@id":"https://ndlsearch.ndl.go.jp/books/R000000004-I028919009"},{"@id":"https://www.jstage.jst.go.jp/article/jnlp/25/2/25_223/_pdf"}],"availableAt":"2018-03-15","foaf:topic":[{"@id":"https://cir.nii.ac.jp/all?q=%E3%83%86%E3%82%AD%E3%82%B9%E3%83%88%E5%B9%B3%E6%98%93%E5%8C%96","dc:title":"テキスト平易化"},{"@id":"https://cir.nii.ac.jp/all?q=%E7%96%91%E4%BC%BC%E3%83%91%E3%83%A9%E3%83%AC%E3%83%AB%E3%82%B3%E3%83%BC%E3%83%91%E3%82%B9","dc:title":"疑似パラレルコーパス"},{"@id":"https://cir.nii.ac.jp/all?q=%E3%83%AA%E3%83%BC%E3%83%80%E3%83%93%E3%83%AA%E3%83%86%E3%82%A3%E6%8E%A8%E5%AE%9A","dc:title":"リーダビリティ推定"},{"@id":"https://cir.nii.ac.jp/all?q=%E6%96%87%E3%82%A2%E3%83%A9%E3%82%A4%E3%83%A1%E3%83%B3%E3%83%88","dc:title":"文アライメント"},{"@id":"https://cir.nii.ac.jp/all?q=Text%20Simplification","dc:title":"Text Simplification"},{"@id":"https://cir.nii.ac.jp/all?q=Pseudo-parallel%20Corpus","dc:title":"Pseudo-parallel Corpus"},{"@id":"https://cir.nii.ac.jp/all?q=Readability%20Assessment","dc:title":"Readability Assessment"},{"@id":"https://cir.nii.ac.jp/all?q=Sentence%20Alignment","dc:title":"Sentence Alignment"}],"relatedProduct":[{"@id":"https://cir.nii.ac.jp/crid/1360011145435838592","@type":"Article","relationType":["references"],"jpcoar:relatedTitle":[{"@value":"The Hungarian method for the assignment problem"}]},{"@id":"https://cir.nii.ac.jp/crid/1360292619788752896","@type":"Article","relationType":["references"],"jpcoar:relatedTitle":[{"@value":"Optimizing Statistical Machine Translation for Text Simplification"}]},{"@id":"https://cir.nii.ac.jp/crid/1360292620352536320","@type":"Article","relationType":["references"],"jpcoar:relatedTitle":[{"@value":"A new readability yardstick."}]},{"@id":"https://cir.nii.ac.jp/crid/1360574093682852864","@type":"Article","relationType":["references"],"jpcoar:relatedTitle":[{"@value":"Proceedings of the Tenth Workshop on Statistical Machine Translation"}]},{"@id":"https://cir.nii.ac.jp/crid/1360574093884296192","@type":"Article","relationType":["references"],"jpcoar:relatedTitle":[{"@value":"Problems in Current Text Simplification Research: New Data Can Help"}]},{"@id":"https://cir.nii.ac.jp/crid/1360574095760310272","@type":"Article","relationType":["references"],"jpcoar:relatedTitle":[{"@value":"Findings of the 2014 Workshop on Statistical Machine Translation"}]},{"@id":"https://cir.nii.ac.jp/crid/1360855569739099264","@type":"Article","relationType":["references"],"jpcoar:relatedTitle":[{"@value":"Text Simplification as Tree Labeling"}]},{"@id":"https://cir.nii.ac.jp/crid/1361137045707694464","@type":"Article","relationType":["references"],"jpcoar:relatedTitle":[{"@value":"Aligning Sentences from Standard Wikipedia to Simple Wikipedia"}]},{"@id":"https://cir.nii.ac.jp/crid/1361418519320955008","@type":"Article","relationType":["references"],"jpcoar:relatedTitle":[{"@value":"SemEval-2015 Task 2: Semantic Textual Similarity, English, Spanish and Pilot on Interpretability"}]},{"@id":"https://cir.nii.ac.jp/crid/1361418521367043072","@type":"Article","relationType":["references"],"jpcoar:relatedTitle":[{"@value":"Minimum error rate training in statistical machine translation"}]},{"@id":"https://cir.nii.ac.jp/crid/1361699994457458432","@type":"Article","relationType":["references"],"jpcoar:relatedTitle":[{"@value":"Unsupervised Sparse Vector Densification for Short Text Similarity"}]},{"@id":"https://cir.nii.ac.jp/crid/1362262944128476928","@type":"Article","relationType":["references"],"jpcoar:relatedTitle":[{"@value":"PPDB 2.0: Better paraphrase ranking, fine-grained entailment relations, word embeddings, and style classification"}]},{"@id":"https://cir.nii.ac.jp/crid/1362262944756694912","@type":"Article","relationType":["references"],"jpcoar:relatedTitle":[{"@value":"Simple PPDB: A Paraphrase Database for Simplification"}]},{"@id":"https://cir.nii.ac.jp/crid/1362544418572213376","@type":"Article","relationType":["references"],"jpcoar:relatedTitle":[{"@value":"Translating from Complex to Simplified Sentences"}]},{"@id":"https://cir.nii.ac.jp/crid/1362544419092988672","@type":"Article","relationType":["references"],"jpcoar:relatedTitle":[{"@value":"A Deeper Exploration of the Standard PB-SMT Approach to Text Simplification and its Evaluation"}]},{"@id":"https://cir.nii.ac.jp/crid/1363107369268322560","@type":"Article","relationType":["references"],"jpcoar:relatedTitle":[{"@value":"Findings of the 2016 Conference on Machine Translation"}]},{"@id":"https://cir.nii.ac.jp/crid/1363107369421120512","@type":"Article","relationType":["references"],"jpcoar:relatedTitle":[{"@value":"SemEval-2014 Task 10: Multilingual Semantic Textual Similarity"}]},{"@id":"https://cir.nii.ac.jp/crid/1363107369929236352","@type":"Article","relationType":["references"],"jpcoar:relatedTitle":[{"@value":"Moses"}]},{"@id":"https://cir.nii.ac.jp/crid/1363388844213567232","@type":"Article","relationType":["references"],"jpcoar:relatedTitle":[{"@value":"Samsung: Align-and-Differentiate Approach to Semantic Textual Similarity"}]},{"@id":"https://cir.nii.ac.jp/crid/1363670320254642688","@type":"Article","relationType":["references"],"jpcoar:relatedTitle":[{"@value":"DLS$@$CU: Sentence Similarity from Word Alignment and Semantic Vector Composition"}]},{"@id":"https://cir.nii.ac.jp/crid/1364233268590612992","@type":"Article","relationType":["references"],"jpcoar:relatedTitle":[{"@value":"ExB Themis: Extensive Feature Extraction from Word Alignments for Semantic Textual Similarity"}]},{"@id":"https://cir.nii.ac.jp/crid/1364233269013320576","@type":"Article","relationType":["references"],"jpcoar:relatedTitle":[{"@value":"A metric for distributions with applications to image databases"}]},{"@id":"https://cir.nii.ac.jp/crid/1364233270826877952","@type":"Article","relationType":["references"],"jpcoar:relatedTitle":[{"@value":"Derivation of New Readability Formulas (Automated Readability Index, Fog Count and Flesch Reading Ease Formula) for Navy Enlisted Personnel"}]}],"dataSourceIdentifier":[{"@type":"JALC","@value":"oai:japanlinkcenter.org:2004293877"},{"@type":"NDL_SEARCH","@value":"oai:ndlsearch.ndl.go.jp:R000000004-I028919009"},{"@type":"CROSSREF","@value":"10.5715/jnlp.25.223"},{"@type":"CIA","@value":"130007397051"},{"@type":"OPENAIRE","@value":"doi_dedup___::a1415a73e601dcfb8589d294a473111b"}]}