Natural Language Processing Technologies for Formulaic Expressions

• Project/Area Number: 18H03297
• Research Category: Grant-in-Aid for Scientific Research (B)
• Allocation Type: Single-year Grants
• Section: 一般
• Review Section: Basic Section 61030:Intelligent informatics-related
• Research Institution: National Institute of Informatics
• Principal Investigator: Aizawa Akiko 国立情報学研究所, コンテンツ科学研究系, 教授 (90222447)
• Project Period (FY): 2018-04-01 – 2021-03-31
• Project Status: Completed (Fiscal Year 2020)
• Budget Amount: ¥17,160,000 (Direct Cost: ¥13,200,000、Indirect Cost: ¥3,960,000); Fiscal Year 2020: ¥5,460,000 (Direct Cost: ¥4,200,000、Indirect Cost: ¥1,260,000); Fiscal Year 2019: ¥5,980,000 (Direct Cost: ¥4,600,000、Indirect Cost: ¥1,380,000); Fiscal Year 2018: ¥5,720,000 (Direct Cost: ¥4,400,000、Indirect Cost: ¥1,320,000)
• Keywords: 定型表現 / 執筆支援 / ドメイン・コーパス / 辞書自動構築 / 意味表現

Outline of Final Research Achievements

This research aims to develop natural language processing technologies to construct a formulaic expression database for English writing assistance. Focusing on academic paper writing, we proposed a method for extracting and retrieving formulaic expressions with their communicative functions. We also constructed an annotated corpus of sentences with communicative functions that can be used to training and evaluating. Our research demonstrated the effectiveness of deep learning models in extracting communicative functions and grammatical knowledge in extracting formulaic expressions.

Academic Significance and Societal Importance of the Research Achievements

英語による文書作成を支援するための定型表現集が数多く出版されているが、その大半は電子化されておらず、電子化されていたとしても執筆途中に必ずしも気軽に利用できるものではない。本研究では、体系化された定型表現を大規模なドメイン・コーパスに対応付けることによって、ドメインに特化した言い回しや伝達機能などを含む豊富な文脈情報が獲得できることを示した。本研究で得られた知見は、これまで困難であった定型表現の予測や検索の実現に結び付くものである。

Research Products

• [Journal Article] Communicative-Function-Based Sentence Classification for Construction of an Academic Formulaic Expression Database. (2021)
  • Author(s): Kenichi Iwatsuki and Akiko Aizawa
  • Journal Title: The 16th Conference of the European Chapter of the Association for Computational Linguistics (EACL 2021): Main Volume.. Volume: - Pages: 3476-3497
  • Peer Reviewed / Open Access
• [Journal Article] Extraction of Formulaic Expressions from Scientific Papers. (2021)
  • Author(s): Kenichi Iwatsuki and Akiko Aizawa
  • Journal Title: The AAAI-21 Workshop on Scientific Document Understanding. Volume: -
  • Peer Reviewed / Open Access
• [Journal Article] A Linguistic Analysis of Visually Grounded Dialogues Based on Spatial Expressions. (2020)
  • Author(s): Takuma Udagawa, Takato Yamazaki, and Akiko Aizawa
  • Journal Title: EMNLP 2020 (2020 Conference on Empirical Methods in Natural Language Processing). Volume: - Pages: 750-765
  • DOI: 10.18653/v1/2020.findings-emnlp.67
  • Peer Reviewed / Open Access
• [Journal Article] An Evaluation Dataset for Identifying Communicative Functions of Sentences in English Scholarly Papers (2020)
  • Author(s): Kenichi Iwatsuki, Florian Boudin, and Akiko Aizawa
  • Journal Title: The 12th International Conference on Language Resources and Evaluation (LREC 2020) Volume: - Pages: 1712-1720
  • Peer Reviewed / Open Access / Int'l Joint Research
• [Journal Article] Assessing the Benchmarking Capacity of Machine Reading Comprehension Datasets. (2020)
  • Author(s): Saku Sugawara, Pontus Stenetorp, Kentaro Inui, Akiko Aizawa
  • Journal Title: The 34th AAAI Conference on Artificial Intelligence (AAAI-20) Volume: -
  • Peer Reviewed / Open Access / Int'l Joint Research
• [Journal Article] An Annotated Corpus of Reference Resolution for Interpreting Common Grounding. (2020)
  • Author(s): Takuma Udagawa, Akiko Aizawa
  • Journal Title: The 34th AAAI Conference on Artificial Intelligence (AAAI-20) Volume: -
  • Peer Reviewed / Open Access
• [Journal Article] Unsupervised Rewriter for Multi-Sentence Compression. (2019)
  • Author(s): Yang Zhao, Xiayu Shen, Wei Bi, Akiko Aizawa
  • Journal Title: The 57th Annual Meeting of the Association for Computational Linguistics (ACL 2019) Volume: - Pages: 2235-2240
  • DOI: 10.18653/v1/p19-1216
  • Peer Reviewed / Open Access / Int'l Joint Research
• [Presentation] タスク指向対話におけるニューラル句抽出を用いた End-to-End 発話生成 (2020)
  • Author(s): 山崎 天, 相澤 彰子
  • Organizer: 第11回対話システムシンポジウム
• [Presentation] 深くて早い言語理解の実現に向けて. (2019)
  • Author(s): 篠田一聡, 相澤彰子
  • Organizer: NLP若手の会 (YANS) 第14回シンポジウム
• [Presentation] A Natural Language Corpus of Common Grounding under Continuous and Partially-Observable Context. (2019)
  • Author(s): Takuma Udagawa abd Akiko Aizawa
  • Organizer: The Thirty-Third AAAI Conference on Artificial Intelligence (AAAI-19)
  • Int'l Joint Research
• [Presentation] Using Formulaic Expressions in Writing Assistance Systems. (2018)
  • Author(s): Kenichi Iwatsuki and Akiko Aizawa
  • Organizer: The 27th International Conference on Computational Linguistics (COLING 2018)
  • Int'l Joint Research
• [Presentation] A Simple Language Model based Evaluator for Sentence Compression. (2018)
  • Author(s): Yang Zhao, Zhiyuan Luo, and Akiko Aizawa
  • Organizer: The 56th Annual Meeting of the Association for Computational Linguistics (ACL 2018)
  • Int'l Joint Research