||Computer Science, Information Technology, Intelligent informatics
||My study field is natural language pocessing and its applications. I have long been interested in the scope and meaning of letter strings, not only in Japanese but also translation tasks. My main research theme is unknown word recognition such as proper nouns, technical terms, embedded foreign phrases, and teen slang phrases. My research also includes emotion recognition, polarity estimation, machine translation, summarization, and automatic recommendation.
|Research Field(Keyword & Summary)
- (1) Unknown word recognition
The word "unknown" means "not available" in natural language processing, such as the words which are not included in the dictionaries so that computers can not recognize. There are plenty of these kind of words, not only proper nouns but also technical terms, foreign phrases, slang phrases and other newly- created phrases. Recognition of a meaningful unknown string of letters itself is not a difficult task, and my research focuses on the extraction of string data such as the meaning of the phrases, original language, part-of-speech tags and the main usage. The results of these research tasks are expected to improve the accuracy of various tasks.
- (2) Emotion/polarity estimation
Many texts, novels and SNS tweets for instance, show emotional signs. Emotion recognition enables to use these emotional signs for the better understanding of the texts. We may need to concentrate on the polarity rather than emotional signs for some kinds of texts such as reviews and comments for automatic recommendation. Our goal is in emotion/polarity estimation includes a better representation of emotions and polarity, and their changes in the texts.
- A study on the use of a color wheel for the representation of emotions, The 4th Asia Pacific Corpus Linguistics Conference, 2018.
- Stepwise acquisition of verbal collocations considering nest structure, The 5th IIAE International Conference on Intelligent Systems and Image Processing, 2017.
- Visualization of keyword transition for Twitter account recommendation, International Symposium on Multimedia and Communication Technology, 2016.
- Patent search based on automatic investigation of topic-related documents, Pacific Association for Computational Lingustic, 2011.
- Robust Toponym Resolution Based on Surface Statistics, IEICE TRANSACTIONS ON INFORMATION AND SYSTEMS E92D(12), 2009.
- ScoreIlluminator: Automatic Illumination of Orchestra Scores for Readability Improvement, Proceedings of the 2009 International Computer Music Conference, 2009.
- Automatic Country Identification of Area Names Based on Surface Features, The 17th Symposium on Natural Language Processing, 2007.
- Three-valued logic for reasoning from an inconsistent knowledge base, Systems and Computers in Japan 37(14), 2015.
- Reasoning with inconsistent knowledge base, Systems and Computers in Japan 37(3), 2006.
- Area identification of English person names based on statistical information, The 19th Pacific Asia Conference on Language, Information and Computation, 2005.
|Recruitment of research assistant(s)
|Affiliated academic society (Membership type)
||(1) IPSJ (regular member)
(2) IEICE (regular member)
(3) ANLP (regular member)
(4) GSK (regular member)
(5) Mathematical Linguistic Society of Japan (regular member)
|Education Field (Undergraduate level)
||Natural language processing, Database system, Programming language processing
|Education Field (Graduate level)
||Advanced Natural language processing