Research Of Lable Recommendation System Based On The User’s Trust And User’s Tagging Motivation In Social Network

Since the birth of the Internet, It had been rolled into the track of rapid development. Especially after Internet steps into Web2.0, “Information Explosion” has become a major feature in modern time. It is so inconvenient for people to search and select as the emergence of information. Based on these problems, the technology methods like vertical classification system, search engine and recommendation system are consistently put forward and adopted by technicians, for solving such problems. Through analyzing the users’ behavior, the personal recommendation system could build the users’ model and predict their preference so as to recommend useful resource to new ones. And this system is adopted extensively by E-commerce and Web Internet Service and receives good feedback in recent years. In the basis of User Generated Content, the UGC tag system can reflect the users’ interest by tagging, and it plays a valuable media role between users and resources. However, there exists individual difference among users, and they understand and think in the different way, which would result in some problems on tagging, such as ambiguous label. At the same time, lacking of trust is also a problem might reduce the effect of individual recommendation.Combined with domestic and overseas research status, this paper studies about lable system, recommendation system and their model algorithms in present time. And it pays attention on users’ credit and tag motivation for building and perfecting individual tag recommendation system in social network. Though constructing users’ collective credit and distinguishing users’ tag motivation, this paper accomplished the function design of the prototype system.In this paper, the main works are:① State the current social recommended system and social network. Study the features of current major recommended algorithm and tag system. Analyze the difference and shortcut of present tag recommendation system.② In order to solve the problem of insufficient coverage of current social networking Web service recommendation algorithm which based on user’ trust. This paper incorporates direct trust, indirect trust and community trust’s research ideas, meanwhile, this paper do a lot extended research on the calculation about trust.③ Aim at the user’ tagging motivation on tag recommendation system, this paper completes the study of calculation and judges about motivation tendency.Besides, this paper created a model based on Ebbinghaus forgetting function and user’ interest to consummate the algoritm of recommendtioan system.④ According to the research of the recommendation based on users’ trust and users’ tagging motivation, this paper designed a prototype recommendation system. And it describes the function model.