The Prediction of User Topic Interest Based on Tags and Interaction of Users

With the rapid development of the Internet, more and more people use social networks to share information and express their views, which lead to a vigorous growth of information. How to select useful and interesting information for users, that is user topic interest, gets more and more attention. Tag functionality in microblog can get user topic interests easily and achieve information recommendation effectively. To deal with the situation that there are few or no tags for most users in microblog, a user topic interest prediction method based on the tag content and interactive relations of users is proposed. The method utilizes Wikipedia to extend the tags and produce a preliminary estimate of the tags' importance. What's more, interactive relations and propagation analysis of different tags are studied. An extensive empirical study using real Sina Microblog data clearly demonstrates that our method has a better performance in user topic interest prediction.