Large-scale three-way group consensus decision considering individual competition behavior in social networks

In social networks, decision makers of group decision-making (GDM) can have interactive be-haviors with different individual relationships. With large-scale group participation, competitive behaviors resulted from competition relationship of the decision makers represent one of the important influences. In this case, the effects of competitive behaviors also have an impact on con-sensus reaching process (CRP) of GDM. Thus, considering the individual competition behaviors in large-scale group social networks, this paper deeply investigates large-scale group consensus decision with increased complexity of GDM in the framework of three-way decision (TWD). More specifically, we firstly improve the K-L algorithm by considering the frequent connections be-tween leaders and followers as well as few connection in own sub-networks under the condition of large-scale group social networks and partition networks for classifying the social roles of leaders and followers. Then, we design the identification rules of competition relationship with consideration of social roles and interest conflict. With respect to the competition relationship, referring to the decision styles of the dependent, biased and rational decision makers, this pa -per further explores the corresponding competitive behaviors in information interaction of social networks. According to the competitive behaviors, we construct a new optimization consensus model in CRP for the evaluation adjustment of large-scale group TWD and obtain the decision result with Bayesian decision procedure. Finally, we apply our proposed model to analyze an example of product promotion and verify its effectiveness.