This study aims to formulate and solve discrete multicriteria decision making (D-MCDM) problems by utilizing artificial intelligent decision support systems. The major advantage of this approach is that data, functions, many D-MCDM methods, choice rules for methods, and the decision maker's preferences in D-MCDM can be integrated in a logical structure. Besides, based on the modulared D-MCDM method base and data base, the decision maker can flexibly choose suitable methods to solve decision problems. This paper first decomposes D-MCDM problems into alternative-attribute, attribute-criterion, criterion-method-recommendation and choice-method relationships, then transforms these relationships into 'Data', 'Function' and 'Rule' formats of logic-based programs. By following that the typical D-MCDM methods as Dominant method, Lexicographic method, Weighting method, ELECTRE method, TOPSIS method and Method with fuzzy concept are coded in a PROLOG-type language in a consistent format. Some choice rules for these D-MCDM methods are then discussed. Finally, the inference process and the man-machine dialog of this system are analyzed.
- Connection Graph
- Production Systems