Methodology and problems of expert assessment. Part 3: group aggregation of pairwise comparison matrices in conditions of heterogeneity of expert competence

This paper addresses the problem of aggregating the opinions of a group of experts in the assessment of complex objects of expertise characterized by a set of heterogeneous attributes. The necessity of accounting not only for inter-expert differences in competence but also for the factor of individual heterogeneity and variability of expert competence is substantiated. A vector model of expert competence is used, which allows specifying competence coefficients for each expert separately for different aspects or attributes of the object of expertise. Three methods for aggregating individual expert judgments into a group vector of attribute weight coefficients are developed: at the level of normalized eigenvectors of pairwise comparison matrices; at the level of pairwise comparison matrices using cross-competence matrices; and at the level of non-normalized eigenvectors employing Hadamard powers. The proposed methods enable flexible consideration of the heterogeneity of expert competence and provide a formalized procedure for obtaining consistent group assessments. The article is intended for specialists in the fields of systems analysis, expert assessment methods, and decision support

Authors: E. A. Burkov, P. I. Paderno

Direction: Informatics, Computer Technologies And Control

Keywords: expert assessment, group expertise, aggregation, expert competence, object of expertise, pairwise comparison matrices, eigenvector, Hadamard product

---

View full article