Research On Mining Algorithm Of User Interest Model Based IPTV

With the rapid development of the Internet,IPTV(Internet Protocol Television)has become more and more popular in people's lives.IPTV provides users with thousands of video programs and television channels,making people's lives full of fun.However,the various video services also lead to the problem of information overload.It is difficult for users to find their favorite programs from numerous video programs in a short period of time.Therefore,it is significant to help users find video programs they might be interested in quickly and accurately.Personalized recommender system is the most effective tool to solve the problem of information overload,which select the appropriate recommendation algorithm and recommend video to users they might be interested in.This paper focuses on personalized recommendation technology based the user interest model for IPTV.In this paper,we analysis the different characteristics of users,and propose two user interest model for different users,User Interest Model based Tag and User Interest Model based Time and Tag.User Interest Model is towards to the group of people whose viewing habits is not regular.User Interest Model Based Time and Tag is towards to the group of people who have regular viewing habits.This paper mainly focuses on the following points:1)Propose the framework for IPTV personalized recommendation system PTV(Personalized Recommendation System),which provides a customized personalized IPTV boot interface,combined with the features of IPTV.PTV allows users to quickly find interesting video.The functions of each module is also described in detail.2)Propose tag interest accumulating algorithm based on the video criterion.The algorithm fixed the deviation of evaluation of user interest caused by the different video length,combining with the concept of the video length criterion.3)Propose a mining algorithm based on the user's viewing habits.The algorithm can automatically identify the user's days and rest days,based on the behavior patterns of different group of popele who have different viewing habits.Based on the user's viewing history data collected by IPTV platform,the expermental result for the two mining algorithms show the effective.