The dynamics model of public opinion diffusion in online social network

Abstract

Due to the potential applications in social and political science, research on public opinion formation and diffusion have been increasing for a long time. Many researchers have developed numerous dynamics models, and they have used a wide variety of statistical, computational, and mathematical methods to understand the spread of public opinion in online social networks. So far, in some standard public opinion diffusing models, the transition probability from ignorants to spreaders is always treated as a constant. However, from a realistic perspective view, an individual whether or not be infected by the neighbor spreader greatly depends on the trustiness between them. In this paper, we introduced a public opinion diffusion model with variable transition probability parameter, in which the infectious probability was partly defined as an exponent function of the spreaders, and covered probability was also defined as an exponent function of the stiflers. Furthermore, we proposed a public opinion diffusion dynamics model with isolation class for hindering the diffusion of public opinion. Finally, we investigated numerically the behavior of the proposed models. Our findings may offer some useful insights in understanding the underlying dynamics of public opinion in online social networks.