Social Network And Role Analysis Based On Micro-Blog

Today Internet is more than just a place where people find information, but an online platform where people collaborate and communicate. Both the Twitter aboard and the sina weibo in our nation have a large scale of users, a huge online social network generated by the following relationships among their users, discussion over a lot of topics and opinions diffusing quickly across the whole network. Today there are more and more studies about online social networks on micro-blog. Since micro-blog is changing how people get the information and communicate with the world, research of micro-blog is of great value. In this paper we studied influence of a node and community detection. Major work is presented as follows:1、Preparations of micro-blog data. Crawling and storing of the data, pre-processing of tweets used in micro-blog searching task in TREC2011.2、Rank users in Twitter and sina weibo by the number of followers and PageRank to calculate score of role’s influence. Results of two rankings are quite similar, but some zombie users get a much lower score by PageRank.3、We crawled the weibo data about7specific topics and generated7topic networks by considering the actions of retweeting tweets as edges of the networks. Based on these networks we ranked users again and found RetweetRank can find the real influential people in one topic.4、 A new algorithm to find community is proposed based on WAF. Affinity between users are calculated and used to find community and we compare it with other famous algorithms.