Social influence or risk perception? A mathematical model of self-protection infection in network

Social influence and risk perception are important to drive the adoption of self-protective behaviors among people during an epidemic outbreak. It is interesting to know whether people should have more belief in social influence or in risk perception when they are faced with an asymptomatic infection. To explore the problem, we develop an extended epidemic model and apply it in a multilayer network topology. It is found that both social influence and risk perception can increase the density of self-protection and reduce the density of infections, but their impacts on the epidemic threshold are different. Specifically, risk perception is unable to affect the epidemic threshold, while social influence can increase the epidemic threshold only if its strength is larger than its own threshold. However, when the epidemic infectivity is very high and the epidemic is ineradicable by self-protection, increasing belief in social influence may instead lead to more infections. The contrary results not only reveal the different mechanisms of social influence and risk perception in hindering an epidemic, but also provide practical implications on the belief distribution between them.