A consensus-reaching method for large-scale group decision-making based on integrated trust-opinion similarity relationships

Recently, the consensus-reaching process (CRP) of large-scale group decision-making (LSGDM) has attracted much attention. Trusting relationships and similar opinions among decision makers have been proven to exert important but different impacts on large-scale group consensus. We propose a large-scale group consensus -reaching method based on integrated relationships between trust and the similarity of opinions ("opinion sim-ilarity"). First, we propose a directed tuple of trust and opinion similarity to blend the information about trust and opinion similarity among decision makers, and construct an integrated relationship network among decision makers based on trust and opinion similarity. Second, leadership-based network partitioning theory is used to divide the integrated relationship network, and the decision maker with the highest degree of recommendation is selected as the representative of each subnetwork. Considering the strength of the integrated relationships and the similarity of opinions of the subnetworks, we add directed edges between the representative of each sub-network and the rest of the subnetworks, to ensure that leaders exist between all subnetworks, which is effective for reaching group consensus. In an integrated relationship network, decision makers' opinions are evolved using opinion dynamics theory to obtain group consensus results. Finally, a numerical experiment is used to illustrate the feasibility of the proposed method, and comparative analysis demonstrates its advantages. The proposed large-scale group consensus-reaching method based on integrated trust-opinion similarity relationships can significantly reduce the initial changes in the opinions of decision makers and is applicable to solving a variety of practical problems related to large-scale group decision-making.