Opinion dynamics in social networks under the influence of mass media

Mass media and social networks have become the primary channels for people to obtain and share information. The news and information to which people are exposed or that they actively acquire can potentially change their opinions on specific issues. Despite ubiquity, the integrated influence of mass media, social networks and individual characteristics on opinion dynamics is less explored. In this study, we propose a general theoretical framework based on improved voter model for opinion dynamics in social networks, which incorporates media outlets, network structure and individual heterogeneity, and allows us to systematically and analytically study the effect of diverse factors on opinion evolution and final steady state. We find that eventually only the opinions held by media outlets or stubborn agents can survive. The ultimate agents with specific opinions generally consist of four parts: stubborn ones, the agents induced by stubborn ones through network links, the agents directly caused by media outlets, and the agents indirectly induced by the effect of media outlets via network links. The model has abundant dynamic behaviors and can be potentially extended to more complicated but more realistic opinion evolution processes. Our theoretical findings are validated by numerical simulations.