1（Department of Computer Science and Technology，Nanjing University of Aeronautics and Astronautics，Nanjing 211106，China）
2（Collaborative Innovation Center of Novel Software Technology and Industrialization,Nanjing 210023,China）
3（School of Foreign Languages and Literature，Yunnan Normal University，Kunming 650500，China）
Abstract：With the constantly improve of complex network theory，the study of network structure and it′s evolution become more applied value，and link prediction as the research focus of complex network has attracted more and more people′s attention.For the large scale networks，because the information of node attribute is incomplete and hard to obtain，so at present the most algorithms of link prediction are based on similarity index of local information.These algorithms always have the simple calculation and better effect of prediction，so they are more suitable for the largescale network applications.But because they usually just considered the amount of common neighbors and degree of nodes.On several networks it′s hard to get the better predictive effect for different algorithms.In this paper we analysis and compare the different points of the different similarity index algorithms.And through analyzed the based structure models we proposed a new LSCN index algorithm.According the structure similarity between one node and another node′s all neighbors，to predict the link possibility.Through abundant experiments we can find that the effect of prediction has a great promotion in several networks.
2018, Vol. 39 
