Algorithms Of Community Detection And User Recommendation

A convenient and effective platform that rapid spread of information and the sharing communication was provided for emerging social network,As the deepening of virtual network integration and the reality of human society,The process of information could be understanded and participated for the social network of users at any time.All kinds of networks were formed by the digital footprints emerged in the environment of information interaction.The related research in social network community detection and users recommendation has been widely concerned by researchers,Due to relational data of users in the network with potential business and research value.This paper mainly studied the association rule algorithm,application in microblog based on graph overlapping community found and user recommendation algorithm,The specific content and the research results were as follows:1.Since the time complexity and superfluous term of Apriori algorithm are high,Firstly,an association rule based on Bitmap and Hashing Algorithm(BHA)was proposed,and the efficiency of the algorithm were effectively improved.Meanwhile,In the application of user recommendation based on microblog,the defect of users' hierarchy distance which was neglected by similarity recommendation was remedied by introducing trust.Applied BHA in microblog user recommendation based on trust,similarity of out-degree and in-degree.The user data of sina microblog was collected to experiment,The theoretical analysis and simulation results showed that the efficiency of mining rules and validity of friends recommendation are improved availably.2.The network model of community discovery in microblog was proposed based on the similarity of user's attention and labels,Combined CNM algorithm and the model to discovery the microblog community.The problem of overlapping phenomenon of node existing in the test results was solved through deleting isolated link based on conception of the core link.Meanwhile,The relationship in common friends is researched,and the algorithm of users recommendation was proposed based on community discovery,The theoretical analysis and simulation results based on real networks and the microblog network show the accuracy of community discovery,meanwhile the recommendation were improved through combining community detection and user recommendation.