A minimum adjustment consensus framework with compromise limits for social network group decision making under incomplete information

Abstract

In consensus-based social network group decision making (SN-GDM) problems, experts tend to refuse modifications if they exceed the limit of their maximum compromise, which may lead to a failed consensus. However, few consensus models have addressed the compromise limit behaviors in SN-GDM. In this paper, we propose a minimum adjustment consensus framework with compromise limits for SN-GDM under incomplete information. First, a two-stage transformation method is proposed to indirectly estimate the social influence based on the preference orderings of experts, and experts’ weights are obtained from the constructed social network using social network analysis techniques. Then, a nonlinear optimization model based on social influence is developed to complete incomplete preferences. After obtaining the experts’ weights and completed preferences, we present a novel feedback mechanism to facilitate the consensus-reaching. The feedback mechanism establishes a minimum adjustment model with compromise limits to generate recommendations for inconsistent experts. Finally, an illustrative example and a comparative study are conducted to show the validity and advantages of the proposed consensus framework.