P2P Live Streaming System with Content Recommendation Based on Users´ Preference

Recently, with the variety of contents in video streaming services, the demand for system assisting users´ content selection is increasing. P2P live streaming distribution is one of the solutions to reduce load of distributors. In P2P live streaming, users select desired contents with the help of content catalog. Users join content network and then, play back some contents. In traditional streaming system, special servers need to collect information for content recommendation in order to assist interesting content selection from multiple contents. This method is unsuitable for P2P live streaming. This paper proposes new P2P live streaming system that can select and distribute recommended contents on time based on users´ similarity derived by access log of contents in an autonomous and decentralized manner.