Information technology of ranking of expert estimates on the basis of the theory of evidence

Abstract

The article concerns the use of modern theories for the analysis of expert information. The results of our research in this area are presented. The main purpose of research is the development of techniques of analysis and aggregation of group expert estimates to solve the problems of ranking. The use of modern approaches of analysis of expert information allows us to take into consideration the forms of interaction of expert judgments, and some specific types of uncertainty, in the frameworks of which the experts’ estimates were formed. This article considers the technology of expert information processing on the basis of methods of mathematical theory of evidence, taking into account alternative techniques of evidence combining. The suggested technique allows to identify and analyze and only relevant alternatives or groups of alternatives. It can significantly reduce the time of the examination, and increase the efficiency of obtained expert information. We suggest the use of the technology for the formation of the collective opinion of an expert group in the case of inaccurate and incomplete expert information. The considered approach suggests new possibilities for the creation of information technologies of expert estimates structuring under conditions of uncertainty, which can be the basis of an automated decision-making support system