In this paper, we first introduce a mutual influence function among network nodes based on characteristics of information spreading in online social network. Then we put forward an information spreading model based on relative weight, analyze the propagation path and process of the network, and discuss the influence on different paths. Finally, the simulation experiments of the traditional SIR model and the model in this paper are conducted with six different network topologies. Results show that the two models have no significant difference in homogeneous networks, but there are significant differences in inhomogeneous networks. This result shows that the information spreading is influenced by the status of spreading nodes, and also shows that the real networks like Twitter and Sina Microblog have certain similarity in topological structure.