An Estimation Framework of Node Contribution Based on Diffusion Information

Zhijian Zhang, Ling Liu, Kun Yue, Weiyi Liu

Abstract: As a key problem in social network studies, identifying important nodes is useful for spreading prediction and restraint. Nowadays, many researches focus on identifying important nodes based on network structure, which is completely necessary. However, the role of a node in diffusion is simultaneously determined by network structure and diffusion characteristics. In this paper, we aim to find the contributive nodes that play important roles in influence spreading without network structure information. First, we formulize the concept of node contribution to influence diffusion to describe the importance of nodes in the spreading processes. Then, we propose an estimation framework and give the method to estimate node contribution based on diffusion samples. Accordingly, the Contribution Estimation algorithm is proposed upon the framework. Finally, we implement our algorithm and test the efficiency on two weighted social networks.