Modeling and Identifying Susceptible Edge Nodes in Echo Chamber Propagation in Highly Aggregated Social Networks

Abstract

Echo chamber affection would block information diffusion and cut down the success rate of delivery. It is of high value to model and identify the user nodes which are the edges of community in social networks as well as the susceptible ones in cascade propagation, especially for the scenario conducting viral marketing and targeted advertising. In this paper, we are interested in the affection of high clustering of social networks to information targeted diffusion and introduce the concepts of susceptible edge nodes to guide the propagation to the gate of echo chamber. We propose a node typing model and use node opinion position evolution as well as structure analyzing and propagation prediction to locate the susceptible edge nodes. Furthermore, the model effectiveness is verified by simulation experiments in this paper. These results are of use in understanding echo chamber propagation and improving the targeted diffusion in highly aggregated social networks.