Influence spreading model for partial breakthrough effects on complex networks

Abstract

Breakthrough effects are a crucial part of many kinds of influence spreading, such as social or infectious contagion. We introduce a novel model that can accurately simulate influence spreading on complex networks with partial breakthrough happening at a given probability. The novel model unifies our earlier analytical and simulation versions of the model that are only applicable to a fixed-breakthrough scenario. A wide range of applications in, for example, social influence and epidemic spreading analysis are enabled by the ability to consider partial breakthrough effects. The breakthrough effects of the new model are controlled by an arbitrary breakthrough probability that determines how likely it is for a node to get reinfluenced. We demonstrate our model on real-world social network structures and provide an example application in the study of epidemic spreading.