I am a research scientist at IBM Research - Tokyo , Japan, working on language and speech processing.
Previously, I obtained my Ph.D. from the University of Tokyo under the supervision of Prof. Akiko Aizawa. I worked on designing dialogue tasks to evaluate, analyze and improve conversational models w.r.t. common grounding. You can find my dissertation here.
[CV] [Google Scholar] [Semantic Scholar]
Journal Papers
Takuma Udagawa and Akiko Aizawa, “Maintaining Common Ground in Dynamic Environments”, Transactions of the Association for Computational Linguistics (TACL 2021), Volume 9, 2021. [paper] [data]
International Conferences
Takuma Udagawa, Yang Zhao, Hiroshi Kanayama, Bishwaranjan Bhattacharjee, “Bias Analysis and Mitigation through Protected Attribute Detection and Regard Classification”, Accepted to the 2025 Conference on Empirical Methods in Natural Language Processing: Findings (EMNLP-Findings 2025). [paper].
Takuma Udagawa, Masayuki Suzuki, Masayasu Muraoka, Gakuto Kurata, “Robust ASR Error Correction with Conservative Data Filtering”, Accepted to the 2024 Conference on Empirical Methods in Natural Language Processing: Industry Track (EMNLP 2024 Industry Track). [paper]
Bishwaranjan Bhattacharjee, Aashka Trivedi, Masayasu Muraoka, Muthukumaran Ramasubramanian, Takuma Udagawa, Iksha Gurung, Rong Zhang, Bharath Dandala, Rahul Ramachandran, Manil Maskey, Kayleen Bugbee, Mike Little, Elizabeth Fancher, Lauren Sanders, Sylvain Costes, Sergi Blanco-Cuaresma, Kelly Lockhart, Thomas Allen, Felix Grazes, Megan Ansdel, Alberto Accomazzi, Yousef El-Kurdi, Davis Wertheimer, Birgit Pfitzmann, Cesar Berrospi Ramis, Michele Dolfi, Rafael Teixeira de Lima, Panos Vegenas, S Karthik Mukkavilli, Peter Staar, Sanaz Vahidinia, Ryan McGranaghan, Armin Mehrabian, Tsendgar Lee, “INDUS: Effective and Efficient Language Models for Scientific Applications”, Accepted to the 2024 Conference on Empirical Methods in Natural Language Processing: Industry Track (EMNLP 2024 Industry Track). [paper]
Takuma Udagawa, Masayuki Suzuki, Gakuto Kurata, Masayasu Muraoka, George Saon, “Multiple Representation Transfer from Large Language Models to End-to-End ASR Systems”, In Proceedings of IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP 2024), April 2024. [paper]
Takuma Udagawa, Aashka Trivedi, Michele Merler, Bhatta Bhattacharjee, “A Comparative Analysis of Task-Agnostic Distillation Methods for Compressing Transformer Language Models”, In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing: Industry Track (EMNLP 2023 Industry Track), Dec 2023. [paper]
Takuma Udagawa, Hiroshi Kanayama, Issei Yoshida, “Sentence Identification with BOS and EOS Label Combinations”, In Findings of the European Chapter of the Association for Computational Linguistics (EACL-Findings 2023), May 2023. [paper]
Takuma Udagawa, Masayuki Suzuki, Gakuto Kurata, Nobuyasu Itoh, and George Saon, “Effect and Analysis of Large-scale Language Model Rescoring on Competitive ASR Systems”, In Proceedings of the 23rd INTERSPEECH Conference (INTERSPEECH 2022), Sep 2022. [paper]
Takuma Udagawa, Takato Yamazaki, and Akiko Aizawa, “A Linguistic Analysis of Visually Grounded Dialogues Based on Spatial Expressions”, In Findings of the Association for Computational Linguistics: EMNLP 2020 (EMNLP-Findings 2020), Nov 2020, pp.750–765. [paper] [data]
Takuma Udagawa and Akiko Aizawa, “An Annotated Corpus of Reference Resolution for Interpreting Common Grounding”, In Proceedings of the 34th AAAI Conference on Artificial Intelligence (AAAI-20), Feb 2020, pp.9081–9089. [paper] [poster] [data]
Takuma Udagawa and Akiko Aizawa, “A Natural Language Corpus of Common Grounding under Continuous and Partially-Observable Context”, In Proceedings of the 33rd AAAI Conference on Artificial Intelligence (AAAI-19), Feb 2019, pp.7120–7127. [paper] [poster] [data]
Preprints
Aashka Trivedi, Takuma Udagawa, Michele Merler, Rameswar Panda, Yousef El-Kurdi, Bishwaranjan Bhattacharjee, “Neural Architecture Search for Effective Teacher-Student Knowledge Transfer in Language Models”, 2023. [paper]
Domestic Conferences
宇田川拓真, 金山博, 吉田一星, “文頭・文末予測の組み合わせによる文特定”, 言語処理学会第29回年次大会 (NLP), 沖縄, 2023年3月. [paper]
宇田川拓真, 相澤彰子, “動的な環境における基盤化タスク設計の試み”, 言語処理学会第27回年次大会 (NLP), オンライン, 2021年3月. [paper] [poster]
宇田川拓真, 相澤彰子, “連続的かつ部分観測的コンテクストにおける基盤化対話コーパスの構築と分析”, 人工知能学会 音声・言語理解と対話処理研究会第84回研究会 第9回対話システムシンポジウム(SLUD-84), 早稲田, 2018年11月. [paper] [poster]
Best Student Award
National Institute of Informatics, Mar 2021.
Committee Special Award
The 27th Annual Meeting of the Association for Natural Language Processing (NLP2021), Mar 2021.
Young Researcher Encouragement Award
JSAI (The Japanese Society for Artificial Intelligence) 9th Dialogue System Symposium (SIG-SLUD 84), Nov 2018.
Oct 2018 -- Sep 2021, Ph.D., Department of Computer Science, University of Tokyo
Dissertation: A Study on Advanced Common Grounding in Natural Language Dialogue Systems [paper] [slides]
Oct 2016 -- Sep 2018, M.Sc., Department of Computer Science, University of Tokyo
Apr 2011 -- Mar 2015, B.Sc., Department of Information Science, Tokyo Institute of Technology
[first_name].[last_name](at)ibm.com
EMNLP 2020 (Secondary), NAACL 2021, ACL-IJCNLP 2021