Considering dual-followers to dynamically model and analyze online rumor propagation

Abstract

Online rumor propagation poses significant challenges to social stability and public security. Previous studies have overlooked the behavioral differences and impacts between active and passive spreaders in the rumor propagation. To address this gap, we propose a novel Dual-Followers–Susceptible–Observer–Infective–Counter–Recovery model (DF-SIOCR), which simultaneously incorporates the followers of rumor-spreading and those of rumor-countering, providing a more fine-grained framework for analyzing online rumor propagation. In the model analysis, we calculate the basic reproduction number, three equilibrium points of the new model, and analyze the stability of these equilibrium points. In the experiments, we analyze the influence of several key parameters on online rumor propagation and conduct simulations on three different network structures to validate the model’s efficacy across diverse social networks. Finally, the comparative experiments are conducted on five public rumor events. Experimental results show that the DF-SIOCR model exhibits superior performance in R-squared, RMSE, and MAE, significantly outperforming existing approaches. These results indicate the model’s high accuracy in predicting rumor propagation trends and strong adaptability to complex dissemination scenarios. This work not only advances theoretical understanding of rumor propagation dynamics but also provides actionable insights for developing effective rumor governance strategies.