The Research And Application Of Domain Community On Social Network

In recent years, with the rapid development of social networking services, user numbers are exploding. Through social networking services, people in addition to conduct the daily social behavior, but treat it a public media platform. The survey found that, in addition to keeping in touch with friends, people use social networks mostly in the acquisition of professional knowledge sharing and track events of interest or topic. In social networks, people have a clear community interaction, users within the same community with the same interest or more focus and communicate closely and connected through different communities associated node.Also, because many users of social networks generating tens of thousands of information every day, for the individual, it is difficult to find a valid concern own content from the mass of data, so we need to study the right way to help users more efficient use of social networks.Against this background, this paper studies the social networking community to identify problems and areas of expertise in the field of professional user community topics monitoring issues. Firstly, the establishment constructs a social network domain community detection model, for the user demand for professional domain knowledge on social networks, the model makes full use of social networks on the basis of data and presentes domain expertise users detection algorithm which can accurately identify areas of expertise in the field of professional expert. Based on the recognized expertise users, this paper completes social network of these users and assess the connection strength of expert users, and community detection algorithm is proposed based on the user connection strength. Then, we build a topic monitoring model of domain community, as for the user can not effectively inform domain experts topic in the face of massive amounts of data generate, based on the full analysis of social network data characteristics and distribution of topics, the paper proposes supervised hierarchical latent dirichlet allocation algorithm, and gives a distributed solution that can efficiently monitor the topcics of domain experts. After verifying real data show that the two models compared to existing solutions, has better performance advantage.Finally, based on researches on social network domain community detection model and topic monitoring model of domain experts, we build a social network domain community topic monitoring system. This paper introduces overall architecture of the system, the results of each module design, development environments and operating platforms, results and performance analysis of system in detail.