Under the epidemic of Covid-19, it is urgent to change from the traditional model of offline experiments to the hybrid model of online-online experiments in computer science needs to be changed from the combination of traditional offline experiments in computer science. To improve the experimental teaching of the database principle in computer science, our teaching team tries to design and build a database system teaching experiment online-offline hybrid intelligence platform. By combing students' active participation and platforms' intelligence, this platform can provide various services such as intelligent recommendation and cocurriculum, enabling to enhance students' understanding of database interaction and cultivate their innovative practice ability. Moreover, in this platform, students can learn the theoretical knowledge of the course through intelligent recommendation for choosing UOOC or online teaching. In addition, students can answer questions to teachers in class online, and complete relevant practices in the platform according to the teaching requirements. Finally, students can perform flipped course display. The application of this teaching platform can improve the quality of classroom teaching and enhance students' practical interest. Moreover, it also can improve students' practical ability and learning creativity, and realize a new teaching mode with students as the main body.