题目: Some Issues in Intelligent Processing of Linguistic Information
报告人： Prof. Hiroay Fujisaki, University of Tokyo
时间： 2014年10月24日，星期五，上午10:00 – 11:30
地点: 北京交通大学 九教北307B
In the first part of this talk, I will take up four aspects of the human language that have traditionally been not very well studied by linguists, but have to be seriously attacked if one wishes to accomplish an intelligent processing of linguistic information. Namely, (1) Ambiguity, (2) Vagueness, (3) Variability, and (4) Openness. For each of these aspects, I will give my own definitions, describe my approaches of analysis, and explain some of the results obtained.
In the second part of this talk, I will give two points that have to be further studied if one wishes to achieve still higher levels of performance. Most of current systems for natural language/speech processing are constructed and operated without paying attention to the way how humans do the task. I will emphasize the need for learning more deeply from the human processes, especially the processes of comprehension and acquisition.
Finally, I will present my view on the ultimate goal of language/speech processing: to facilitate human communication, processing, and sharing of information. I will conclude this talk by emphasizing the fact that the purpose of language/speech processing by machine is to facilitate transfer of information, knowledge, and even ways of thinking from one human mind to another, which will become possible only if one has a model of the human mind and of the interaction between two or more individual minds.
Hiroya Fujisaki Professor Emeritus The University of TokyoResearch Areas
Speech Communication and Spoken Language Processing, Natural Language Processing, Human and Artificial Intelligence, with special emphasis on modeling, including a model of language use; a model for perceptual processes in speech identification/discrimination (known as the dual channel model); a model for the process of fundamental frequency control in speech (known as the Fujisaki model); but also, a model of road traffic flow and its application to road traffic control systems.
Education and Professional Career: BS, MS, PhD from University of Tokyo; Fulbright Scholar at MIT ('58 -'61), Guest Researcher at KTH ('60) Joined the Faculty of Engineering, University of Tokyo ('62), Full Professor at EE Dept. ('73 - '91), serving twice as Chairman, Also Professor of Speech Science, Graduate School of Medicine ('74 -'77) and Lecturer, Dept. of Linguistics, Graduate School of Humanities ('64-'91)
Prof. Emeritus, University of Tokyo ('91- present); Professor, Tokyo University of Science ('91- '06)
Visiting Professorship: U. Texas, Austin, U.S.('80); KTH, Stockholm, Sweden ('81), U. Goettingen, Germany ('81),.Nanjing University, Nanjing, China ('85-'86), University of Science and Technology of China, Hefei, China ('87-'91).
Project Leader, the Japanese National Project on Advanced Human-Machine Interface Through Spoken Language ('87-'91)
Project Leader, the Japanese 'Research-for-the-Future' Project on Human-Machine Spoken Dialogue Systems ('96-'00)
General Chairman, IEEE-ASJ International Conference on Acoustics, Speech, and Signal Processing (ICASSP-86) ('86)
ASJ-Chairman, the Second Joint Meeting of the Acoustical Society of America and the Acoustical Society of Japan ('88)
Founder and Chairman, the International Conference on Spoken Language Processing (ICSLP) ('90)
Member, the Permanent Council for the Organization of ICPhS ('79 - '07)
Life Member, the Permanent Council for the Organization of ICSLP ('92-present)
Member, ISCA Board ('00-'07), Life Member, International Advisory Council of ISCA ('04- )
President, the MIT Association of Japan ('82 - '90)
President, the Boston Association of Japan ('97 -'99)
Vice President, the Japan-China Science and Technology Exchange Association ('03 - present)