There is an extraordinary amount of video content available online now through popular websites like YouTube, Netflix, and Hulu. With the rise in the number of online streaming video websites, it has become increasingly challenging for users to quickly identify video content that falls in line with their interests.  That challenge has given rise to the recommendation system. Traditional recommendation systems, which employ collaborative filtering, predict the interests of users by mining their video watching and rating history data. The popular online social networks, like Facebook and Twitter, provide additional information to improve the accuracy of video recommendation. The common rationale behind social recommendation systems is that a user’s taste is similar to and/or influenced by her trusted friends in social networks. NYU WIRELESS faculty member Prof. Yong Liu is developing a Facebook app for music video recommendation between friends. The core of the app is a social recommendation system that automatically generates music video playlists for users by mining their own video watching and rating history, as well as video watching and rating activities of their friends on a social network.  User feedback, including likes and dislikes, comments, and shares, are collected and processed in real-time to dynamically update their playlists. If the app is widely adopted by users, it will serve as a valuable platform to conduct online experiments to test and improve the recommendation accuracy and user experience of various social recommendation systems proposed by other research teams.

Folllow Research