The blog world is a representative online society. To understand the nature of the blog world, there have been many research efforts on analyzing information diffusion and blog-ger activities. The independent cascade model is appropriate to explain information diffusion in the blog world. For the model to be employed, the blog world should be represented as a form of a network. For accurate analysis, it is crucial to assign a diffusion probability to each edge between a pair of bloggers in the blog network. In this paper, we propose a novel method to assign a diffusion probability to an edge for a pair of bloggers that reflects well the phenomenon of actual information diffusion between them. We verify the superiority of our approach by performing extensive experiments with real-world blog data.