ISSUES IN KNOWLEDGE REPRESENTATION AND REASONING

Timo Honkela
Helsinki University of Technology
Adaptive Informatics Research Center

Invited talk at N.S.R.C. "Demokritos", Athens, Greece
6th of June, 2006

Abstract

Knowledge representation and reasoning are important areas of artificial intelligence. They have been considered as the basis for areas like natural language processing and expert systems. Traditionally,knowledge representation and reasoning have been based on certain analytical means, such as first order predicate logic and related formalisms. Human beings, however, reason in a probabilistic manner and reasoning is highly contextualised by relevant prior knowledge and belief. It seems that a division into a heuristic system and an analytical system can be made. The heuristic system has evolved early, it is shared with animals, it is rapid and parallel, has high capacity and is pragmatic. Therefore, considering the analytical level is not enough as knowledge is based on the underlying experiential domain. In this presentation, I will consider the methodology for developing adaptive systems, knowledge representation and reasoning from various points of view. In particular, I will discuss the implications on semantic web and natural language processing technologies.