A new method of defining divergent opinions in group decision-making

Owing to different personal preferences, knowledge backgrounds and life experience of decision experts, divergent opinions always exist in group decision-making, and influence the assembly of the final opinions in varying degrees. Therefore, the determination and treatment of divergent opinions appear to be extremely important. For the existing divergent opinions in group decision-making, this paper proposes a new method. Firstly, in each of the alternatives, we calculate expert closeness degree matrix (the matrix consists of all the close degree between any two expert opinions) based on the concept of closeness degree in fuzzy theory. Secondly, in the light of the extent of disagreement, we take different treatment methods accordingly. Finally, the processed group opinions are synthesized by the evidence combination formula. The new algorithm defines the divergent indexes and divergent experts in order to modify the divergent opinions pointedly, and it is implemented in MATLAB, reducing the amount of computation and the waste of resources.