Information Propagation Dynamic Model Based on Rumors, Antirumors, Prom-Rumors, and the Dynamic Game

The propagation of Internet rumors has complex dynamics. The traditional propagation model, based mainly on rumor and antirumor messages, provides a static analysis of the rumor propagation situation. This article introduces the concept of a prom-rumor (promoting rumor) message and thoroughly investigates competition and cooperation among multitype rumor messages. First, we employ a multiple linear regression method to construct the information influence mechanism by extracting network structure, topic content information, and historical behavior information, considering the complexity of spreading multitype compound rumor messages. Then, we introduce game theory and formulate a game strategy to capture the competition and cooperation among rumors, antirumors, and prom-rumors. Furthermore, we consider the interactivity in multitype compound rumor messages and combine the dynamic game to construct the driving mechanism of rumors, antirumors, and prom-rumors. Finally, we define the user-state transfer equation and propose a dynamic model of information propagation based on rumors, antirumors, prom-rumors, and the dynamic game while considering the multidimensionality and polymorphism of multitype compound rumor messages and combining mean-field theory and the infectious disease model. In validation experiments on real-life data, the model accurately captures the game propagation process of the multitype rumors and predicts the propagation trends in rumor networks.