User Interest Modeling For Mobile We Media Based On Community Discovery

In recent years, mobile Internet is developed in the true populist wave of information dissemination. In The Times of everyone to participate in the Media.We Media is unprecedented popular. More and more people are involved in information dissemination activities to share their own things and the comment the content that their friend published. Since we media gradually enter into people’s lives, people’s interaction through the social network platform is more and more frequent, producing a large number of user data, and fill in every field and aspect of people’s life with enormous size. At the same time, due to user data is updated faster, some valuable information may be lost. People need to spend a lot of time and energy to get the interested content. In this case, in order to satisfy their interests(content recommendation, recommend friends, recommend video, and other forms), intelligent recommendation and personalization services have arised.Both intelligent recommendation service and personalized retrieval service, its core is to build the user interest model, existing methods of building user interest model are mostly interested in user’s data mining, getting users interest categories. This kind of model has the advantages of reflect the user’s interest accurately, but user is affected by other members of the community. In order to solve this problem, this paper use complex network theory to study the mobile social network community features, through mining the interest user generated content and found the interest community in the mobile social network to build the user interest model. Processes are as follows:In the first place, the paper consider the complex network structure, and combine the theory of complex network with the characteristics of the mobile social network analysis, giving the method of establishing user relations of complex network model. Then, the paper makes a deep study of the community discovery algorithms. Aiming at the shortcomings of the existing algorithms, the paper propose a community discovery algorithm based on "prototype community" of complex network combining with the characteristics of the mobile social network. The algorithm used in several classical network in social network has been tested. Experiment shows that algorithm has higher accuracy in the community division results and can be used to build the user interest model.On this basis, considering the diversified characteristics of users, the paper thinks that the user interest is influenced by the other members of the community. then combined with user personal interest model comes the final user interest model present. Finally this modeling method has carried on the experimental analysis, the evaluation index found that the method has good performance in interest findings.