Information Random Propagation Behavior in Social Networks Considering Node Centrality

Abstract

In order to be close to the real characteristics of information propagation in social networks, a random propagation model considering node centrality is proposed. The experiment on Facebook dataset was carried out based on Monte Carlo simulation method. The experimental results show that under the same propagation probability, the propagation speed is the slowest when the max clustering conefficient node is the initial central node. The two information propagation curves with the max betweenness and the max degree as the initial central node are highly consistent, and the information propagation speed is obviously faster than the max clustering conefficient of the initial central node. The research results provides theoretical basis for predicting and analysising the development of network public opinion information.