Knowledge acquisition is the foremost procedure in knowledge based systems. It is the procedure whereby non codified human expertise is transferred into a well codified computer program (G. Buchanan and C. Wilkins, 1993). Principally, it involves two major issues. Firstly, the accessing of human expertise from domain experts, and secondly, the converting, or the modeling of this expertise into a `pre chosen' form in a computer language, known as knowledge representation, for a system. Many successful knowledge representations and knowledge acquisition techniques and tools have been developed to make this procedure as automatic and accurate as possible. However, problems still exist in these well defined and developed representation and tools. Many such problems are related to the fact that they are non object oriented (OO) programs and the knowledge representations used in the programs are non object oriented knowledge representations. By discussing the good features of existing knowledge representations, knowledge objects and OO knowledge representation are described. Based on this knowledge representation and the analysis of the knowledge acquisition process, an interactive OO knowledge acquisition approach is also designed
Published in:
Knowledge and Data Engineering Exchange Workshop, 1997. Proceedings
Date of Conference: 4 Nov 1997