Personalized Information Recommendation On The Microblog Media

Microblog media has a large number of users, which generate a lot ofUGC(User Generating Content) information. People communicate with each otheron the microblog media, publish, share and disseminate information. People need tospend a lot of time in order to find interested information. Therefore, it isparticularly important to help people to find their favorite information.Personalized information recommendation collects and analyzes user’sinformation to learn their interest and behavior in order to achieve the purpose ofinformation recommendation on microblog media. This paper designed anddiscussed how to recommended microblog information that comply with user’spreferences. It’s on the basis of building a user’s interest model, and analyzing thecorrelation of the microblog information with the model.Firstly, this paper analyzed the characteristics of the data on microblog media,and given the method for getting the user’s information. Then the characteristics ofrepresentatives of user’s interest extracted from the user’s information and given aunified microblog media user’s interest modeling methods.Secondly, this paperproposes a group of information recommendation methods, which includes twoaspects:1) The quality of the information itself;2) The correlation of theinformation with the user’s interest model. The second aspect contains there factors:1) Content characteristics;2) Personal behavioral characteristics;3) Groupbehavioral characteristics. And we also give out the related experiments shows thatthe quality of the information itself and the personal behavioral characteristics areimportant than other factors in information recommendation on the microblogmedia. Finally, this paper give an introduction of the microblog application basedon the work of build the user’s interest model and information recommendation.The application’s name is “User’s interest analysis system on microblog media”.