Abstract

As digital information environments become increasingly dynamic, the spread of misinformation presents a fundamental challenge to content governance, public trust, and knowledge-based systems. Traditional techniques, such as rule-based fact-checking and manual content moderation, are often inadequate for real-time, high-volume digital platforms. This chapter introduces agent-based modelling (ABM) as a core computational approach for simulating the complex, adaptive, and often non-linear behaviours that underpin infodemics. The chapter begins by highlighting the limitations of conventional statistical models in capturing the stochastic and exploratory tendencies of human information-seeking behaviour. It then outlines how ABM provides a bottom-up simulation architecture in which heterogeneous agents interact within rule-governed environments to produce emergent belief dynamics and social effects. The use of network topologies to represent information diffusion, and the incorporation of reinforcement learning (RL) algorithms to model adaptive agent behaviour, allow for the simulation of belief transitions, echo chambers, and misinformation cascades in a way that mirrors real-world complexity. This chapter sets the foundation for thinking about infodemics not as isolated data points, but as complex systems, where agent diversity, uncertainty, and feedback loops must be accounted for in both analysis and intervention. Through ABM and RL, Computational Infodemiology moves from observation to simulation, enabling the testing of strategies for misinformation mitigation in silico before real-world implementation.