A topic-oriented analysis of information diffusion in a blogosphere

Abstract

A blogosphere is a representative online social network established through blog users and their relationships. Understanding information diffusion is very important in developing successful business strategies for a blogosphere. In this paper, we discuss how to predict information diffusion in a blogosphere. Documents diffused over a blogosphere deal with information on different topics in reality. However, previous studies of information diffusion did not consider the information topic in analysis, which leads to low accuracy in predictions. In this paper, we propose a topic-oriented model to accurately predict information diffusion in a blogosphere. We also define four primary factors associated with topic-oriented diffusion, and propose a method to assign a diffusion probability between blog users for each topic by using regression analysis based on these four factors. Finally, we show the effectiveness of the proposed model through a series of experiments.