The Research Of User Influence For Social Network

With the rapid rise of the Internet,people are more inclined to get all kinds of information they need on the network,and ways to communicate with others transfer from the traditional off-line to on-line.A variety of social network platforms have sprung up on the Internet.On the social network platform,users can easily share their experience,concern about their interest in current events and meet like-minded friends.The number of users of social networks grows like index.Hundreds of millions of Internet users are producing vast amounts of content data every day,and these kinds of content information is spread to every corner of network by a large number of users,influencing the behavior of other users.Diffusion of information relies mainly on the user's behavior and user's influence in the network.The greater the user's influence,its information dissemination capacity is stronger.This paper studies the influence of social network users,mainly divided into two study sites:1.Assessment of the influence of social network usersThe study found that users often has a relatively large influence in some areas they are good at,and they mainly concern the information in these areas.So it is more meaningful to assess the influence of users in specific areas.Firstly,based on field division,author assess the influence of users in their field.Based on relationship between users and content user generated,this paper presents a comprehensive method to assess the influence of user,and the experiments show that the method effectively eliminates some cheating factors for user influence assessment on the network.Finally,the user influence applied to predict user relationship and microblog retweeting,and preliminary prediction model is proposed.2.Influence maximization problem of social networkThe target of influence maximization problem is to select a specified number of seed node from the network,so that the information is spread most widely in the network via seeds nodes.The study found that users of social network are more prone to have emotional resonance and interactive information with similar users;the more active users,its greater ability to disseminate information;users are more inclined to accept information from their usual friends.In this paper,according to user similarity,user activity and user relationship,information dissemination model based on social network(IDM-SN)is proposed.According to community of social network and connectivity and information diffusion capacity of users,influence maximization algorithm based on community connectivity(IMA-CC)is proposed based on this model.The experiments show that IMA-CC algorithms have achieved good results at aspect of time efficiency and information dissemination coverage.