In the context of mobile Internet, the art teaching platform pays more attention to anytime, anywhere, and users can learn anytime, anywhere, regardless of the location of the environment. Therefore, it is necessary to design a reasonable interactive teaching system for analysis and testing, provide users with matching and learning of online art courses, and provide corresponding response measures. To properly solve the problem of course matching and selection, we have developed a new system that combines matching trees and embedded technology. Firstly, the matching tree algorithm is used for similarity diagnosis between art courses and user preferences and has achieved good results in similarity diagnosis and matching. Secondly, an interactive teaching system architecture for online art courses was built using embedded technology, which improved the efficiency of the interactive teaching system. Finally, the applicability of the established similarity matching model was verified through simulation testing of the learning system.