Abstract:Inspired by practical situations in which individuals must make group decisions by grouping, we propose an extended Hegselmann-Krause model to describe the opinion dynamics over a clustered social network with intermittent communication. First, a hierarchical clustered network was built where the agent with the largest degree in each cluster is defined as the opinion leader. Second, an intermittent opinion dynamics model was developed using a bounded confidence rule. Each time interval was divided into two consecutive sub-intervals. During the first sub-interval, the agents only communicated with those in the same cluster, and during the second sub-interval, only the opinion leaders communicated among the clusters. Finally, a quantitative analysis was performed for the simulation results, which showed an interesting opinion-wave phenomenon: the cluster leaders' opinions fluctuate during the different stages of communication before the consensus appears.

Opinion Formation over Clustered Social Networks with Intermittent Communication