Using The Analytic Hierarchy Process in Frame-Based Expert Systems

Takao Terano
1988 unpublished
Expert systems techniques in Artificial Intelligence give an effective way to implement sophisticated software systems. One of the difficulty to apply expert systems techniques to decision making is caused by the fact that some of the knowledge intrinsically has uncertain information. This paper proposes a new method which uses the Analytic Hierarchy Process (All?) for handling uncertain knowledge in frame-based expert systems for classification type decision making problems. The method
more » ... The method consists of the following two phases: i) Construct a "part-or frame-taxonomy for the target problem; Then give consistent scores to each item in the frame by applying the AHP. Describe empirical rules; The rules first evaluate the scores in the frame, then derive a plausible conclusion. The proposed method gives some "rational" measures by means of the Consistency Index or the Consistency Ratio to uncertain knowledge, and furthermore it improves the limitation of the AliP i.e., the difficulty to construct a proper hierarchy and the simple decision criteria based on the additive value functions.
doi:10.13033/isahp.y1988.086 fatcat:ylgk77pmwzdqne4hso6tvanxii