Model-Based and Data-Driven Stochastic Hybrid Control for Rumor Propagation in Dual-Layer Network

Xiaojing Zhong,Chaolong Luo,Feiqi Deng,Guiyun Liu,Chunlei Li,Zhipei Hu
DOI: https://doi.org/10.1109/tcss.2024.3469972
2024-01-01
IEEE Transactions on Computational Social Systems
Abstract:Toward exploring the positive impact of media debunking and random blocking on the spread of rumors, we discuss a stochastic hybrid control strategy that combines an individual and media debunking method, a continuous stochastic blocking method, and an impulse interruption method. Using stochastic analysis, the almost sure exponential stability of the controlled system is analyzed, along with the expression of control intensities. To balance rumor suppression, minimize control costs, and enhance the generality of control, a data-driven machine learning (ML) approach is developed to provide suboptimal control solutions. Numerical simulations based on two real-case datasets are carried out to validate the theoretical results and evaluate the potential impact of the model-based, data-driven stochastic hybrid control strategy.
What problem does this paper attempt to address?