Design of expert system for alternative selection in multi-attribute decision making

Abstract

This research suggests an expert system for alternative selection within the domain of Multi-attribute Decision Making (MADM). This system can be used whether attribute weights exist or not. Since obtaining specific weights of attribute as input are stressful and time-consuming for users, friendly decision-aid system is suggested. Especially, partial information with behaviorally consistent set. (BCS) is allowed. However, if conflict occurs in the constraint set of weights, the conflict resolution system resolvs the inconsistency to produce modified BCS. Based on the modified BCS, the alternative ranking or best alternative can be obtained. This system is installed at personal computer using Insight 2 which is shell for rule-based expert system.