Latent Dirichlet Allocation Hash Tagging Approach for Friend Recommendation in Microblogging System

Microblogging is becoming people's most attractive choice for getting the information and expressing opinions because of the developing universality and small frame. Messages got by a user mainly rely on whom user follows. Therefore, recommending user with related interest may enhance the experience quality for information receiving. Since messages posted by Microblogging users reflect their hobbies or interest and the important keywords in the messages show their primary focus to a huge extent, we can find users' interest by investigating the user generated contents. Besides, user's hobbies, interest are not static; despite what might be required, they change as time changes. In light of such instincts, we proposed a LDA model in microblogging system for friend recommendation to analyze user's possible behavior’s and predict their potential friends in Microblogging. The model takes into users’ potential preferences by extracting keywords from aggregated messages over a period of time using a topic model, and after that, the effect of time is considered to deal interest.