In the blog network, the posts in a blog can be diffused to other blogs through trackbacks and scraps. Analyzing information diffusion in the blog network is an important research issue that can be used for predicting information diffusion, detecting abnormality, marketing, and revitalizing the blog world. Existing studies on information diffusion in a blog network define explicit relationships between blogs and analyze the word-of-mouth effect through such explicit relationships only. However, it has been observed that more than 85% of all information diffusion in a blog network occurs through non-explicit relationships. In this paper, we propose a new model that considers both the explicit and non-explicit relationships between blogs in order to explain the information diffusion phenomena in a blog network. We add a super node and the relationships between the super node and blogs as broadcast edges and register edges to the existing information diffusion model and assign the assimilation probability to every relationship. The expanded information diffusion model improves the accuracy of the basic model by taking into account the degrees of diffusion powers of posts. We verify the superiority of the proposed model through extensive experiments of information diffusion at a real blog network. The experimental results show that our expanded information diffusion model generates 77% less errors than the existing model.