Discrete time information diffusion in online social networks: micro and macro perspectives.

Opinions shared publicly in online social networks spread broadly and at an extremely high speed. However, modelling information diffusion in online social networks is still a challenge that is intriguing to many researchers. To monitor public opinions online, it is necessary to model the process of information dissemination. In this paper, we first study information diffusion based on the network structure and time occupation. By taking into consideration the availability of a user, e.g., his online or offline state, we present the discrete-time bi-probability independent cascade model. We next analyse the information diffusion from a macro perspective. A diffusion model is established by merging the interferences from other events and the cumulative effect that occurs over time. Finally, we observe the factors in online social networks that impact a message's diffusion from a micro perspective and discuss more complex user behaviour and various types of interferences with their effects from a macro perspective. Experiments are conducted with real world data, and the experimental results justify our models.