Rating prediction algorithm and recommendation based on user beahavior in IPTV

摘要: Service quality of IPTV directly influence Quality user's Experience (QoE), one the key technologies to attract new users. The current researches mainly focus on two aspects: On hand, researchers are concerned evaluation videos; other personalized recommendation is cared more and more. For former, most effective solution improve bandwidth network; but second, Collaborative Filtering (CF) Algorithm performs perfect effect in service. This paper we pay attention later, based interests user. Owing characteristic interactions between user television platform, different behaviors user, such as explicitly rating behavior, watching behavior saving so on, may show Items. To obtain make Personal recommendation, author firstly introduced related mining algorithm according main three then proposed a similarity computation CF. Finally performance evaluated with modified data from real TV provided by Wenguang Shanghai Corp. China it shows quite comparative recommendations.