Discovery Method Of Micro-blog Community Based On H-Index

In recent years, micro-blog are becoming the core of social network. After having a separate service platform, it are gradually evolved into a new form of information dissemination. Currently, the number of micro-blog registered users in China has exceeded5billion. There are huge potential valuable knowledge to mine from the micro-blog platform.In process of the micro-blog network, part of user will gradually form a small group. This group is a kind of social community phenomenon. If we can mine the small group users in which have same or similar interests, it will be useful to help users to focus the concern, and at the same time, we will hava a more efficient because of precision advertising.In order to meet demand on the goal of micro-blog marketing of infant products, this paper proposes a method of micro-blog community discovery based on user influence. By using the method of zombie fans recognition based on user behavior model, the experiment excludes lots of zombie fans in a user community, to ensure the purity of the community. The community discovery method this paper proposed combines with user influence ranking based on user influence, text classifier based on support vector machine, recognition method of zombie fans based on user behavior and so on. Through the real data acquisition, the experiments analysis the results from different dimensions.