Modeling and simulation of rumor propagation in social networks based on Petri net theory

In this paper, we remodel the rumor propagation process using Petri net and demonstrate the possibility of describing dynamic behaviors of social networks in Petri net. First, we transform popular Susceptible-Infected-Removed epidemic model into Petri net, from which single rumor model is derived. The single rumor model denotes a basic unit of information propagation and can be tuned for other kinds of information like clarification. By merging rumor and clarification units and making them compete on tokens, the model for multiple pieces of information propagating is finally built up. According to rumor spreading data collected from Weibo.com, we primarily tune the parameters of the model. The model output is compared with real world data and explanatory power of the model is achieved.