A group consensus method based on social network and three-way decision under multi-scale information systems

This paper synthesizes and analyzes relevant data at multiple levels (i.e., multi-scale information systems (MSIS)) by fusing social network (SN) and three-way decision (TWD). It enables us to effectively address the complexity and uncertainty intrinsic to rich decision making environments. In addition, in group decision making (GDM) it is likely, often desirable, that agents view problems from different perspectives. Therefore, they need to reach consensus through negotiated discussions and changes in opinions. Such discussions will stop when either the group reaches a satisfactory consensus, or its members' willingness to adjust reaches a maximum value. Within this context, the goal of this article is to investigate the consensus reaching process (CRP) of decision makers in an MSIS setting, and to establish a least-cost consensus method by fusing SN and TWD perspectives, referred to as the CRP-SN-MSIS method. First, optimal scale combinations in MSISs are filtered based on the spirit of TWD. In this way, we obtain multiple decision makers from a GDM viewpoint. Then, a measure method of their trust relations is designed to establish a social network among decision makers. Subsequently, the detailed process of the CRP-SN-MSIS method is presented and applied to a case study targeting a real dataset. Finally, the applicability and superiority of the designed method is fully validated through both qualitative and quantitative arguments.