Knowledge extraction changes the way an expert thinks

The authors show an aspect of knowledge extraction that deserves attention: the process where a human domain expert becomes more cogent, in his or her own mind, about the algorithms, heuristics, and strategies that he or she uses in manipulating this factual knowledge base. The contention is that the knowledge engineer's constructs can be greatly strengthened by utilizing those constructs of the domain expert that may evolve through the knowledge extraction process. In addition, harnessing the domain expert's constructs by fostering such introspective processes early in the knowledge extraction stage will increase the cost effectiveness of the expert system design. The arguments are based on a case study of one particular human expert who has been involved from 1985 to 1992, in the construction of five expert systems, all of which pertain to the same domain and have the same objective to assist in the diagnosis of stuttering among young children.>