The Social Influence Analysis Based On The Static Network Structure And The Dynamic Spread

With the rapid development of information technology, social networking carries a large amount of data and information as people online dating and the important network life platform in a social network information transmission, user friend recommendation, Internet viral marketing, experts found, advertising, public security and other fields has been widely applied. Social influence analysis is an important part of the study of social networks.In this paper, we can study the analysis model and algorithm from the static social network structure and the dynamic propagation process influence, respectively for social network influence measurement method. Introduces the basic concept of social networks, social networks influence of related factors, and a variety of measurement methods, describes the influence maximization propagation model and algorithm, etc. The main content of the thesis are as follows.First of all, this paper analyzes respectively the influence measure and spread biggest problem based on the structure of network, and then introduces the research status quo of social influence, describes the basic concepts of social network and common analysis methods and analysis tools. This paper introduces also the origin, definition and related factors of social influence.Secondly, put forward the concept of behavior activity and the similarity and calculation method based on user behavior influence analysis insocial network. Aiming at the problems of existing influence measurement methods, BASR algorithm was proposed based on user behavior characteristics, making social influence more objective and accurate.Then, definition and analysis for the most influential transmission problem in social network. This paper introduces the two kinds of the influence of the traditional communication model, elaborates the influence maximization algorithm of traditional calculation method in detaile and relevant concepts aiming at proposing a hybrid optimization greedy algorithm. To improve the node activation metrics and greedy strategy respectively, so that the algorithm is more accurate and efficient.Finally, we can carry on the experiment on the standard data sets for the two algorithms proposed in this paper respectively in the social network, and accuracy ratio and the number of active node as evaluation index of two algorithms is analyzed.