Modeling and simulation analysis of public opinion polarization in a dynamic network environment

摘要

At present, there are many group incidents on the Internet, which has caused a high level of public opinion. Hot issues on the Internet can often trigger intense discussions among netizens and finally evolve toward bipolarization or multipolarization. This article focuses on the phenomenon of public opinion polarization under the dynamic network. First, the single-dimensional evaluation system for the JA model was expanded to multidimensional and analyzes the main factors that cause the polarization of public opinion of network groups from multiple dimensions of hot events. Second, the strength of the relationship between individuals and its dynamic changes were taken into account in the model. The strength of the relationship between individuals and its changes will affect the interaction of opinions, and thus affect the degree and speed of group polarization. Third, combining the changing process of the strength of relationship, the dynamic connection mechanism of network nodes was set to make the network structure more consistent with the social network in real life. Finally, we verify the reliability and scientificity of the model by combining practical cases.