Research Of IPTV Recommendation System

Expecting a more intelligent TV? With numerous alternative living channel,users have to spend most of the time on selecting what they really like.To help users quickly locate their favorite channels,an effective recommendation system which is capable to predict user behavior is urgently needed.The main application scenarios of IPTV recommendation are EPG recommendation and preload recommendation.This paper firstly introduces current research status and related techniques of IPTV recommendation system.And then,we design a recommendation strategy for traffic burst channel,including data preprocessing and preparation,passive listening mode,active pull mode,switching state judgment.The main application scenario of this strategy is EPG recommendation.In the passive monitoring mode,traffic burst channels can be added to the recommended list,and the recommended list would be filled according to the channel supplementary rules in the active pull mode.Meanwhile,a recommended strategy for concurrent channels streaming is designed and a channel recommendation framework based on popularity is proposed.The problems of traffic growth and the long switching time are also discussed in this chapter.The main application scenario of this strategy is preload recommendation.By modeling the situation when the user requests arriving server,the strategy builds a CCS channel selection framework based on popularity ranking,creates an economical and practical buffer method,and discusses how many channels should be included in the CCS.The implementation of the strategy and related experiments based on user behavior are presented in the chapter five and an IPTV recommendation system is designed.In conclusion,we design two IPTV channel recommendation strategies in this paper which reduce the occupancy of server resources and the consumption of network bandwidth to provide a better user experience.