Twitter is one of the largest social network services, and it involves only s single social link of follow. However, most of users follow other users in different communities, in which a same tweet is seemed to be distributed different ways. Therefore, information propagation in twitter can be considered as overlaps of single layers of communities. If it is true, we can estimate an information propagation extent of a tweet from the combination of information propagation models for the single layers extracted from actual tweets. In this paper, we assume that a single layer can be represented by a weighted adjacency matrix and the combination of the single layers is modeled as the addition of the matrices. Then, the objective of this paper is to develop a method to represent the matrices and the combination and verify the accuracy of the model. We investigate the method with actual tweets and show it is valid enough.