Kiyoaki Shirai Associate Professor
School of Information Science、Entertainment Technology Area
B.S., M.S and Ph.D. from Tokyo Institute of Technology
Associate at Tokyo Institute of Technology(1998)
natural language processing, machine learning
corpus based natural language processing
Knowledge acquisition for natural language processing
Various types of knowledge resources such as dictionaries or thesauriare required for sophisticated natural language processing.We aim at obtaining such knowledge resources automaticallyfrom corpora and/or machine readable dictionary.We also aim at enriching existing language resources,i.e. enlarging them or adding new information to them.
WWW access support using natural language processing technology
While World Wide Web contains vast information, obtaining desired information from WWW often entails too much labor. Our goal is supporting people to access WWW easily using natural language processing technology. For example, we currently aim at automatically constructing a portal site, developing the system to monitor when and how web pages are updated, and so on.
Statistical Natural Language Analysis
Disambiguation is one of the major problems in natural languageanalysis. This research aims at statistical disambiguation, i.e.analyze sentences using various types of statistics learned from verylarge corpora.
- Acquisition of Anecdote from Web for Free Conversation System，Kiyoaki Shirai, Tomotaka Fukuoka，Asean Japan Joint-Workshop on Computational Linguistics & Informatics (AJCLI 2017) / Asia-Pacific Journal of Information Technology and Multimedia，6，2，13-22
- Re-examination of Training of Semantic Class Disambiguation Classifier for All Words，Truong Vo Huu Thien, Kiyoaki Shirai，The Eighth International Conference on Knowledge and Systems Engineering (KSE)，187-192，2016/10/7
- Recognition of Sarcasm in Microblogging Based on Sentiment Analysis and Coherence Identification，Piyoros Tungthamthiti, Kiyoaki Shirai, Masnizah Mohd，自然言語処理，23，5，383-405，2016/12/15
◇Lectures and Presentations
- SENSEVAL-2 Japanese Dictionary Task，Kiyoaki Shirai，SENSEVAL-2: Second International Workshop on Evaluating Word Sense Disambiguation Systems，2001/7/5
◇Academic Society Affiliations
- The Institute of Electronics, Information and Communication Engineers，Member，2015-
- The Japanese Society for Artificial Intelligence，member，2001-
- The Association for Natural Language Processing，member，1998-
- The 24th annual meeting of Natural Language Processing Society in Japan (2018)，Program committee member
- The 23rd annual meeting of Natural Language Processing Society in Japan (2017)，Program committee member
- The Association for Natural Language Processing，Director，2012/04/01 - 2016/03/31