Blog Community Discovery Based On Social Network Analysis

Blog as a social network, is a relatively young social learning and communication tool. it is increasingly becoming an important source of information. Community mining is a relatively new research directions. Community structure and content analysis has important practical value. With the development of the network, which carries information about getting rich, how to dig out from these valuable, meaningful information content is particularly important, that communities found provide a good way.Community found mainly divides the net into a number of designated groups, in order to accelerate the formation of spontaneously gathered Blog user community. Traditional community method mainly analyzed linsk between the different node structure and text content, using a layered approach to divide the community,and the community structure is usually rendered tree.Due to the rapid development of the Internet, the traditional algorithm based on a single characteristic does not make for a new social networking site. With the development of graph theory and data mining technology, many clustering methods have also been used to discover classification blog community.In this thesis,,firstly, we analyses a single individual blog, its structure and characteristics were analyzed.secondly, we focus on the social network research, including of analyzing the behavior of individual characteristics and connection features.giving a method of community found based on text content and link relationships.Then in terms of text Extraction,we analyze mechanical theme extraction and semantic grammar points system. Found in the technical aspects of the community, at home and abroad to study summarize and compare their advantages and disadvantages, social network analysis based on this consideration, and to analyze the relationship between the text and links Blog users find communities. The information mining of content and comments are based on LDA. And then design related experiments. The results show that we can get the potential community and topics.Finally, given the results of the work,analysed some problems may exist in the future and gave more reasonable suggestions for improvement. In this paper, after a lot of research and experiments, we get the direction of the future research Blog community discovery needs to try, such as to identify better way to use model to analyze the text content of Blog community,community discovery algorithm for further optimization.