Channel recommendation method for high-burstiness user request

Abstract

The invention discloses a channel recommendation method for a high-burstiness user request. The method comprises the following steps: firstly creating a recommendation list and an alternative list byobtaining basic data (comprising user viewing history records and current data); then, when a user views a channel, starting a passive monitoring mode in the channel recommendation method, selecting channels meeting conditions, and respectively pushing the channels to the alternative list and the recommendation list; when the user enters a startup trigger time interval of an active pull mode, starting the active pull mode, and updating and filling the recommendation list; and finally, after the user completes the channel switching, judging whether to update the recommendation list and the alternative list, and re-entering the monitoring mode. The method can be applied to IP TV electronic program guide EPG recommendation and IP TV channel preloading, response can be made to the high-burstiness user request quickly to automatically perform intelligent channel recommendation, the operation is stable and reliable, and the time cost and the space cost are both relatively low.

Description

technical field [0001] The invention relates to the technical fields of IP television, Internet video and smart television, in particular to a method for channel recommendation oriented to highly sudden user requests. Background technique [0002] The continuous emergence of various smart terminals provides a richer platform for users to watch video-on-demand and live channels. Among them, the emergence of IP TV brings better video viewing experience for users and brings new challenges for operators. . With the emergence and popularization of IP TV, video content operators are gradually increasing, and the channels and content that users can watch are also increasing. Not only the program content of the channel shows a trend of diversification, but also the attributes of the channel are also based on the fact that most of the comprehensive channels in the past On the Internet, many personalized channels have been derived, and the needs of the audience have also derived pers...

AI Technical Summary

Problems solved by technology

[0004] The recommendation of live channels is very complex. At present, the algorithms based on personalized recommendation include content-based recommendation, collaborative filtering-based recommendation, machine learning-based recommendation and hybrid recommendation method. Although the above algorithms are used in video recommendation All have good performance, but there are still some problems: 1. The existing recommendation methods have large time or space overhead; 2. The existing recommendation methods are mainly used in video-on-demand systems; 3. Live-oriented The recommendation method cannot respond in time when faced with sudden user access; 4. The recommendation method for live broadcast does not combine real-time recommendation with historical recommendation; 5. The solution for sudden access during live broadcast is not included The stable mechanism makes some channels repeatedly enter and exit the recommendation list, resulting in a waste of resources

Images