Role of propagation thresholds in sentiment-based model of opinion evolution with information diffusion

• Propose a model to study opinion dynamics with information diffusion.
• Two sentiment-based propagation thresholds are introduced.
• Large infected threshold restrains information diffusion and weakens the extremism.
• Large refractory threshold facilitates decision exchange and promotes the extremism.

The degree of sentiment is the key factor for internet users in determining their propagating behaviors, i.e. whether participating in a discussion and whether withdrawing from a discussion. For this end, we introduce two sentiment-based propagation thresholds (i.e. infected threshold and refractory threshold) and propose an interacting model based on the Bayesian updating rules. Our model describe the phenomena that few internet users change their decisions and that someone has drop out of discussion about the topic when some others are just aware of it. Numerical simulations show that, large infected threshold restrains information diffusion but favors the lessening of extremism, while large refractory threshold facilitates decision interaction but promotes the extremism. Making netizens calm down and propagate information sanely can restrain the prevailing of extremism about rumors.