On the Development of an Expert Decision Support System Based on the ELECTRE Methods

Abstract

Analytical justification of decision options using decision support systems (DSS) significantly improves the quality of decisions. The use of the currently existing DSS, which usually includes one or two decision-making methods, does not always lead to the desired results, since each method is based on certain assumptions and is not universal. The noticeable effect is achieved when many decision-making methods are included in one DSS knowledge base. The systems that meet these requirements belong to the class of Expert Decision Support Systems (EDSS), which currently includes more than 50 decision-making methods. Expanding the EDSS knowledge base, made by including new methods in it, allows choosing the most suitable solution method for each decision-making task. Addition of the decision table model, which is the basis of the system knowledge base, allows developing EDSS without completely processing the system program code. The ELECTRE methods were adopted for expanding the EDSS knowledge base. The basis for the selection was their key feature, which consists in the fact that they do not use the alternative valuation convolution operation given in different scales according to individual criteria. The article shows the adapted algorithms of the ELECTRE family methods ready for inclusion in EDSS. The algorithm is proposed for obtaining a criterion matrix being based on alternatives that serve as input information for the ELECTRE methods in cases where there is no objective information to fill it. The results of the study can be used to develop EDSS, that open the way to analytically substantiate solutions using methods that were not previously used in the system.