A novel fuzzy DEMATEL method based on social network analysis and α-level sets considering experts’ hesitant information

The purpose of this paper is to investigate the method of determining experts' weights based on social network analysis (SNA) and to propose a novel fuzzy DEMATEL method based on α- level sets to identify critical factors in supplier evaluation. First, we employ hesitant fuzzy linguistic preference relationships (HFLPRs) to accurately express experts' preferences among evaluation attributes. Subsequently, using SNA, we specify the centrality index of the experts to calculate the expert dependency, which in turn yields the weights of each expert. Next, using the α-level sets, we transform the triangular fuzzy numbers (TFNs) into confidence intervals and constructed three correlation preference matrices via probability density function to maximize the preservation of the original information. Subsequently, we process these three preference matrices by applying the DEMATEL method to identify the key factors under different conditions. Finally, we demonstrate in detail the feasibility and superiority of the proposed method in this paper through case studies and comparative studies.