Steady-state analysis of a three-level compartmental model for disinformation propagation in a set of agents

Abstract

Disinformation has become prevalent in social media risking the flow of vital information especially during crisis. A simple three-compartmental model (DSI) is introduced to model the dynamics of disinformation and information propagation in social media systems. Agents may assume one of the three states: susceptible (S), disinformed (D), or informed (I), which is dependent on the type of information an agent holds and disseminates. The system has two parameters: (1) reverting rate, r, which is the state of transition from either D or I state back to S state, and the (2) learning bias rate, α, which is applicable to interactions induced by an agent in I state. Deriving an equivalent ordinary differential equation system (ODE) that correlates to the model's stoichiometric rate equation allows for the steady-state analysis of DSI. The system is stable when the conditions of the system allows it to be purely informed or purely disinformed.