Rearch On Microblog Content Recommendation Based On User Relationships And Interest

With the development of Internet technology, it is very common for people communicating on the Internet. Surfing the Internet has became part of daily life. Internet provided all kind of services to people while its size is also explosive developed.The internet has a wealth of useful information,but it is hard for people to find useful information from vast amounts of data , people's demand for personalized information has became a serious problem to be solved. Recent years, with the development of social network , people's daily life has extended from the real world to the online world , especially with the appearance of social network such as facebook, twitter, Sina microblog and so on.It has became an important way to get the information from network. With the gradual expansion of the scale of social networks and the quick update of information, the problem of information overload in social networks have emerged.The traditional recommendation system has achieved great success in many areas,such as film and e-commerce. However, social networks have characteristics different from other areas where there are a variety of factors of user's personal interests,such as social relationships , trust relations and so on, so that social networking is more complex on personalized recommendations .With the rise of the social network,users left a big amount of data on social network sites,but the requirement of real-time message made the cold-start problems and data sparsity problems more serious.Based on the users interest and relationships between users,to satisfy the demand for personalized information,we studied two aspect,which are content recommendation based on constraints of socialization and based on the constrains of trust. The main contents and results of this study are summarized as follows:In this paper,we proposed a new method to recommend microblog messages by introducing social factors.This method can recommend messages with cold start characteristics.In social networks,people interested in the microblog messages not only because the user interested in the information itself,but also contains the users'need for social communication.By combine the users' society factors and the micro-blog content information,we can recommended cold start characteristics messages more accurately.We proposed a recommend method based on trust constraints.Introducing trust relationship between users mitigated the sparse problem.In social networking sites,the familiarity between users and other social attributes form the recommended trust between users.In this paper,we defined the direct trust between users is that the more frequent communication between users the stronger trust between them.Trust relationship transfer from one to another which discovered the potential social relationship between users,so that the data sparsity problem can be relieved,and then improved the accuracy of microblog messages recommendation.In this paper, a method of recommend microblog messages were studied. Microblog platform is a popular social networking platform,the information overload problem has appeared with the number of users increases and the growth of datasets. By the in-depth analysis of the microblog platform,and combining with traditional method,we achieved a great result to recommending high-quality microblog messages to users.