Study Of Affinity Based Information Diffusion In Online Social Networks

In recent years, with the emergence of Web2.0as well as the rapid development of online social networks(OSNs), which occurs revolutionary change in the way of people acquiring and disseminating information. The unique characters of information diffusion in OSNs has drawn amount of attention from many scholars. But existing relevant studies have certain limitations due to the unilateral considerations of diffusion factors such as social, information or network structure.In this paper, we propose the concept of information diffusion, which consists of the level of users’interests in information-user interest a[i] and the attractiveness of information to the users-information threshold6, the probability of diffusion information depends on the interaction of a[i] and6. Then we propose an affinity based information diffusion model, and run the simulation and empirical research in both Social Network Site and Microblog Network.The results show that in SNS, the information diffusion range is related to the information threshold θ, the network average degree k, the probability of immunization p2. The probability of questioning can only affect the speed of information diffusion. In microblog network, the node forwarding probabilities of popular microblog diffusion tree in entertainment, finance and social themes are similar with uniform, power-law and normal distribution; in each topic, the average node forwarding probabilities of popular microblog diffusion tree differ with the popularity of the microblog. The information diffusion range is related to the information threshold θ and the active fan probability of node pl.