Nation Chiao Tung University
Hsinchu, 300, Taiwan, R.O.C.
Knowledge acquisition is known to be a critical bottleneck in building expert systems. In the past few years, many researchers working in this field have focused on the development of methodologies and systems to shorten the time needed for eliciting expertise and to improve the quality of knowledge bases. Some reports in this field show that the missing implicit knowledge may be a possible factor which significantly affects the performance of expert systems. In this paper, we extend our previous work on the elicitation and analysis of implicit knowledge from experts to resolve the consistency problem in the process of knowledge elicitation. We analyze each possible inconsistent condition that may arise in eliciting implicit expertise and then propose a consistency test to help experts to cope with the problem.
Keywords: consistency, expert systems, knowledge acquisition, personal construct theory, repertory grid
Received October 7, 1993; revised August 26, 1994.
Communicated by Soo-Chang Pei.
*This original version of this paper was presented in the IEEE Fourth International Conference on Software Engineering and Knowledge Engineering.
This study supported by the National Science Council, Republic of China, under the contract number: NSC82-0113-E-009-295-T.