Research On Community Discovery And Excavation Of Structural Holes For Chinese Microblog

With the rapid development of Internet technology, virtual social networksgradually become the focus of attention. People can more easily make friends inthe virtual community platforms, so as to promote the development of real socialrelations and make the application integration of human activities and theInternet. As a new network platform for information exchange, the microblog'sunique design provides more excellent and convenient communication. Also,large user groups contains vast amounts of text information and great valuewhich is of great significance in different the research field. In this paper, themain research for microblog network as follows:Firstly, we introduce the basic concept of social networks and the mainfeatures and analyze the theory and techniques of network analysis such as thepretreatment of the text, the text representation model and commonly usedclustering algorithms. while the theory of semantic text analysis is researched tounderstand the necessity of feature space dimensionality reduction and the basicflow of latent semantic analysis.Secondly, we analyze the main characteristics and exchange mechanisms formicroblog network and apply the microblog platform to achieve open access todata. Meanwhile, microblog network shows the multi-dimensional structuralfeatures, and the structure of the conventional planar network model does notreflect the degree of authority under a different subject areas. Therefore, wepropose a multi-layered network model based on the theme and establish thedifferent levels user interaction network diagrams of different themes to dig upthe microblog social groups.Thirdly, we analyze the microblog network community features and applythe word co-occurrence of the event and the progress of latent semantic analysisto realize the foundation of topic communities based on the similarity ofcommunity theme and the traditional community discovery algorithms.Meanwhile, this paper proposes a Greedy algorithm based on maximumcardinality search and the WEBA algorithm based on the weight and balance toachieve a reasonable division of the network of community structure from the network structure of communities.Finally, we elaborate the basic concept of structure hole in social network,and the importance of social assets as well as different areas to make animportant evaluation of network nodes. Meanwhile, this paper proposes the HISalgorithms based on the opinion leaders and the MaxD algorithm based onminimum cut set to achieve the mining of structural nodes from the perspectiveof information diffusion and the structure of network community.