Hiroyuki Kameda
In my research laboratory, we have been engaged in a wide range of research fields to clarify still unmeasurable and uncovered human intelligence scientifically, so that we can apply its acquired knowledge to computer software. The followings are research topics we are now interested in, for example:
(1) Blockchain3.0 technology to keep reliability of knowledge that AI systems discover and create automatically from other data and knowledge other computers acquire,
(2) clarifying a computational model of human higher brain functions to create games of tailor-made cognitive rehabilitation programs based on it, and
(3) smart NLP (Natural Language Processing) systems with ability of acquiring unknown words and syntactic rules inductively based on ILP (Inductive Logic Programming) methodology.
Name | Hiroyuki Kameda |
Department/Research Field | Natural Language Processing |
Research content |
Language is a tool for human to communicate with each other, to think, and to describe knowledge with use of it. It is socially invaluable for computers also to have capability to take a command of it. In our research, we are aiming at creating a new sophisticate model of acquiring words and syntactic rules unknown to an AI system inductively and gradually, as if the system were a human child, and at the same time building a smart AI system with ability of acquiring unknown words and syntactic rules from a large scale of text databases. Target languages are Japanese and English, and Prolog programming language is adopted to implement the system. Inductive logic programming system Progol is affirmatively applied to it. |
Figure 3. Smart parser to acquire new grammatical knowledge (unknown words, unknown syntactic rules and etc.) of languages.
|
Research theme |
1. Smart natural understanding system to acquire unknown words and syntactic rules inductively and gradually |
Research keywords |
unknown words acquisition , unknown grammar acquisition , inductive logic programming |
Lab link |