Leader’s opinion priority bounded confidence model for network opinion evolution

Meixia Zhu,Guangqiang Xie

DOI: https://doi.org/10.1063/1.4992877

2017-01-01

AIP Conference Proceedings

Abstract:Aiming at the weight of trust someone given to participate in the interaction in Hegselmann–Krause’s type consensus model is the same and virtual social networks among individuals with different level of education, personal influence, etc. For differences between agents, a novelty bounded confidence model was proposed with leader’s opinion considered priority. Interaction neighbors can be divided into two kinds. The first kind is made up of “opinion leaders” group, another kind is made up of ordinary people. For different groups to give different weights of trust. We also analyzed the related characteristics of the new model under the symmetrical bounded confidence parameters and combined with the classical HK model were analyzed. Simulation experiment results show that no matter the network size and initial view is subject to uniform distribution or discrete distribution. We can control the “opinion-leader” good change the number of views and values, and even improve the convergence speed. Experiment also found that the choice of “opinion leaders” is not the more the better, the model well explain how the “opinion leader” in the process of the evolution of the public opinion play the role of the leader.