In the real world, there are many complex networks with a certain number of nodes. Complex network theory has been used in many fields such as World Trade Web. As an important part of complex network, several social network models have been proposed to simulate the social networks in real world in the last decades. While there are plenty of characteristics and current models cannot describe them. In this paper, a new network model is proposed, whose name is the Interests Drift Network Model (IDNM). A network closed to real-life can be created by IDNM, because the formation of communities is based on nodes' various preference of fields of interest, and each node's interest fields may change every once in a while. In the experiments, we simulate the growing process of IDNM and measure some characteristics. It's shown that there are differences between IDNM and traditional network model: the parameters such as modularity and degree diversion are more similar to real-life networks.