Threshold Aggregation of Fuzzy Data Using Fuzzy Cardinalities of a Set of Fuzzy Estimates

The problem of threshold ranking of alternatives represented by vector fuzzy estimates of a three-gradation scale (high, medium, low) is considered in the article. A generalization of the corresponding procedure for fuzzy estimates is proposed, based on the calculation of the fuzzy cardinalities of a set of fuzzy estimates for each gradation. We discuss the general properties that the fuzzy cardinality of a set of fuzzy estimates must satisfy. The proposed method is illustrated by the example of ranking articles based on their assessments by reviewers and considering information about the degree of confidence of reviewers in their assessments. This information from reviewers is converted into vectors of three-grade fuzzy ratings for each article. These alternatives are then ranked according to the developed methodology. The differences between the results of such ranking and the ranking of unblurred estimates and/or after applying alternative methods are discussed.