Communication Engineering Laboratory

K. Matsuzawa

Kazumitsu Matsuzawa

Office: 23-624 Ext.: 3813


03/2002, Engineering Dr., Tokyo Institute of Technology

03/1977, Engineering M., Tokyo Institute of Technology

H. Amanuma

Hiroshi Amanuma
(Technical Assistant)

Laboratory: 23-621 Ext.: 3810


03/1984, B.Eng., Kanagawa University

Research Field

Communication Engineering (Artificial Intelligence, Natural Language Processing).

Research Overview

Communication engineering helps facilitate communication between people and supports mutual understanding between human and computers. Our research extracts knowledge from the vast amount of information on the Internet to advance smooth collaboration between people with different ideas.

Research Subjects

A system for collecting reliable knowledge from the Internet, studies of mechanisms to evaluate miscellaneous knowledge comprehensively, a method for predicting whether advance story from fragments of conversation, a system that uses the sense of words in the selection of goods, and elucidation of the intelligence required for communication.


To understand human words, computers require data similar to the common sense that human beings have. For example, the content of human languages can be translated, and it is evaluated with the knowledge that different words can point to the same thing. Therefore, we have built a database of the common-sense meanings of words by using the language knowledge on the Internet. This is called the Concept-Base. With the Concept-Base, the degree to which the meaning of certain words resembles other words can be calculated as a number between 0 and 1. By using this calculation (RUIJING), computers will be able to understand human language better.
However, humans are also sensitive to the sound of words. For example, you might associate a word in a conversation with another similar-sounding word, even though the meanings might be different. This association can spread to the main topic of conversation. New ideas can also arise. We created a mechanism to calculate the degree to which the sound of certain words resembles other words as a number between 0 and 1. By using this calculation (DAJARENG), computers can apply human words to various situations.
Furthermore, combining our two systems can better aid communication between humans and between humans and computers. Using these support technologies, communication engineering aims to realize a society where humans and computers can cooperate closely.

  • 1)N. Inago, K. Kasahara, and K. Matsuzawa, “A method for judging the semantic similarity between nouns by using noun-noun co-occurrence in compound words,” Journal of Information Processing Society of Japan, vol. 41(8), pp. 2291–2298 (2000).
  • 2)K. Fujimoto, H. Kazawa, H. Sato, A. Abe, and K. Matsuzawa, “DSIU systems: Decision support for Internet users: Providing hot choice by using information on the Internet,” Journal of The Japanese Society for Artificial Intelligence, vol. 15(1), pp. 261–64 (2000).
Affiliated Academic Organizations

K. Matsuzawa:
The Institute of Electronics, Information and Communication Engineers (IEICE), Information Processing Society of Japan, The Japanese Society for Artificial Intelligence, The Association for Natural Language Processing.

Current members
◯ Professors: 1 ◯ Technical Assistants: 1
◯ Postgraduates: 1 ◯ Undergraduates: 13