Researchon Evaluating Node Influence Insocialnetworks

Along with the development of Web2.0, online social networks are growing rapidly and becoming main platforms for pepole making friends and sharing informationon the Internet.On social networks, users can get to know each other, share and disseminatea variety of information.Massive information generated by huge amount of users on online social networks has enormous commercial value and research value, which attracts more and more researchers to study the social networks from various perspective. Evaluatingnode influence in social networks is one of the hot research direction. By evaluating the influence of node in social networks, high influential people can be mined, which is meaningful for advertising, information control and user action analysis.The main work of this paper is researching the method of evaluating node influence in social networks to find influencial people more effectively. Based on studying the existingmethods of evaluating nodeinfluence, a new method of evaluating node influence in social networks is proposed in this paper. The method is validated experimentally.Now few researches about evaluating node influence in social networks consider the topical feature of influence and user ability of spreading information in different communities. Focusing on these problems, a new approach to evaluate node influence is proposed in this paper. The approach consider users’relationship, users’ posting and forwarding action and posts contents together to measure users’ topical influence. Closeness and social circle difference between users is defined to calculate users’influence rank in different topics. The experiment dataset is from Sina Weibo containing users information and their posts information. Experiments show that our approach can mine different influential users in different topics and the ranks in different topics are little correlated. The influential users also have strong spread ability. The higher user influence is, the more evenly his posts forwarded times in different communities.