Yu. Ulianovskaya

Full text PDF

Annotation. The questions of processing expert information provided by the group of experts are considered. In the treatment group of expert opinion there is a problem obtaining a generalized result. It is assumed that the information is a set of qualitative and quantitative features (alternatives), described by linguistic concepts. Statistical methods of expert’s data processing are quite complicated when the expert’s answers have form of ranking or separation, and quite simple, if the answers are the results of independent pairwise comparisons. In this article are proposed to carry out processing of expert information by methods of the fuzzy sets theory. Using this theory are developed a method for determining the qualifications of the expert based on his length of service and number of expertise conducted by him, the results of which are accurate. The developed method is based on the construction of a system on fuzzy logic with fuzzifikator and defuzzifikator. For processing expert’s estimates are suggested each alternative presented in the linguistic variable form and evaluate it by assigning a group of experts of membership functions each term by the direct method. Obtaining a generalized assessment based on all expert’s estimates going on with regard to their competence. In this paper a method of ranking fuzzy alternatives are propose. Based on the developed methods of data processing was designed an automated system that allows to determine the experts qualification, generalized result of expert group evaluation and of ranking alternatives. The developed technology is applicable to any subject area, where it is necessary to analyze alternatives for many of the criteria based on processing of expert estimations.

Key words: expert information, linguistic variable, fuzzy logic, ordering alternatives, qualification of experts, information technology.

"fuzzy logic"

an international quarterly journal on economics of technology and modelling processes