Research On Social Network User Influence Evaluation Algorithm

With the development of mobile Internet,social network has become an important communication way.Through the social network,people can publish information and have a live chat.The popularity of social network produces a new way of information dissemination.In the change of people's way of life,social network has attracted a lot of scholars to study it.Because of convenience and real-time,Sina weibo,as a new era social networking tool,has attracted a large number of users recent years.According to incomplete statistics,the number of active users of Sina weibo has reached 236 million unitl 2016.Sina weibo has a large number of users and the speed of information transmission and renewal is fast.There is too much information overload problem.Each user has different influence in the network.So we can reduce the information overload and improve user satisfaction by evaluating user influence and evaluating user influence is of great significance.This paper presents a new algorithm for computing the influence of micro-blog users.According to the social network topology,an iterative factor is added in the new algorithm which is based on PageRank algorithm.We can get the value of the user's influence by the method of a number of iterations and the weighted quantization.The main work of this paper is:1.We obtained the true data in the social network and analyzed the number of fans,the number of friends,the absolute rate of attention to each other,the relative rate of attention to each other,and so on.According to complex network and social network theory,we studied the characteristics of social network of nodes,edges and the network topology.Taking Sina weibo as an example,this paper analyzed the formal definition method and message delivery form of Sina microblogging network and provided the theoretical basis for later.2.This paper studied the properties of users.User attributes are divided into fixed and non-fixed attributes.The number of fans and the number of friends which are affected by the fixed property can reflect the influence and activeness of users.This paper made a statistical analysis of user attributes and evaluating user influence indexes are obtained.3.This paper presented a new algorithm for computing the influence of micro-blog users.According to the social network topology,an iterative factor was added in the new algorithm which is based on PageRank algorithm.We can get the value of the user's influence by the method of a number of iterations and the weighted quantization.The experimental results of new algorithm compared with that of PageRank algorithm.