Research On Key Technique Of Social NetWork Analysis

In the Web 1.0 era, information is released by professional Web sites. Users can only browse and download information. And they cannot participate in the production and dissemination of information. Since the 21 st century, Web 2.0 is gradual on the rise and developing rapidly. The mode in which information is produced and propagated is also gradually changed. Users only can browse and download information before, but now they can produce, spread and read information. Users’ role changes from the original information consumer to information consumer and producers. As a result, people will spend a large part of time on the Internet, especially social networks, to study and work. So, social networks have become an essential part of people.On social networks, huge amounts of data is generated every day. By analyzing these data, you can get potential information, such as people’s preferences, geographical distribution, living statues and trends of some importent events. The information has great social values and commercial values. Thus, the government can adjust the policy with the relevant information, and the company can also adjust company strategies based on analysis results, and so on. Therefore, it is of great theoretical and practical significance to study and research on social networks.Based on the Douban network, this paper studies the structural of social networks. Research work has three aspects. First, this paper verifies the validity of basic structural characteristics of social networks, including scale-free, small-world and community structures. Second, nodes are analyzed in the "degree centrality" and "closeness centrality" of social networks in this article. Third, this paper studies the " recursion shingle algorithm", and put forward the improved method "adding environment nodes" on the basis of this algorithm. In the third part, besides research on “recursion shingle algorithm”, through the comparing the “central node set” that was obtained by analyzing the node to the "community node set" was obtained by community mining, this paper discovered that the phenomenon of “one central node needn’t be included in community structures ”, and analyzes and vertifies the cause. On this basis, this paper proposes the social network node classification scheme, and puts forward some suggestions for the social network promotion.