The Impact of Emotion on the Information Diffusion Model in Online Social Network

Abstract

In order to characterize the information propagation accurately on online social network, we propose an optimized SEIR information diffusion model. Firstly, the model considers the emotional factor of every node, which is a dynamic parameter in propagation process. The model defines five possible states: susceptible, exposed, removed, positive and negative infected. Every node attached to a specific emotion value, the dynamic interaction among these emotion values is produced in the diffusion process. We analyse the new mechanism of information dissemination and establish the mean-field equations. As we all know, the information dissemination is affected by user’s emotion, user’s degree and other factors. By simulating information diffusion process on real social Weibo, we analyse the density of users and the effect of some factors in the diffusion process. Simulation results agree well with the theoretical analysis, and it shows that the optimized model fits the characteristics of online social networks.