In the era of mobile internet, the new media video entertainment industry has developed rapidly, and entertainment information is showing an explosive growth trend. The problems of information overload and knowledge redundancy are becoming increasingly severe. How to efficiently and accurately obtain interesting entertainment information has become a difficult problem in people’s daily lives. Recommendation systems are one of the important applications of data mining and an important way to solve information overload. The recommendation system mainly includes login registration module, personal center module, management permission module, video recommendation module, video browsing module, video editing module, and comment and like module. This article analyzes the specific requirements of the video service platform, designs the system architecture and specific modules, and develops and tests the system. After testing, all parts are currently running well and meet user expectations.