Abstract: Through the survey and research of the existing video sites，some video sites face the problem of information overload. So it is necessary for the video sites to have a recommender system. Through the study of the existing video recommender system, using open source cloud computing technologies—Hadoop, and some of its related components such as Hive and Hbase, a personalized video recommendation system based on cloud platform was designed. The system is only applicable to professional video-based website.