Knowledge Acquisition Techniques for Expert Systems: Conceptual and Empirical Comparisons

Abstract

Abstract : The knowledge-acquisition phase in the development of expert systems is hampered by inadequate techniques for the elicitation and representation of knowledge from human experts. The objective of this research is, ultimately, to develop guidelines for effective knowledge acquisition. The objective of this Phase I (SBIR) research is to establish the feasibility of designing and executing experiments comparing knowledge-acquisition techniques. These empirical comparisons should provide the data that permit us to categorize the kinds and amount of different kinds of knowledge that each technique elicits, so that the technique can be matched with the domain of expertise and the purpose of the resulting expert system to provide the knowledge engineer with the most effective means of building the knowledge base. Keywords: Artificial intelligence; Cognitive psychology; Expert systems; Knowledge acquisition; Route planning.