Abstract:  In the context of artificial intelligence and higher education, this study explores the construction and implementation of intelligent courses in university physics. The research is based on knowledge graphs and adopts a hybrid teaching mode that combines online and offline methods, aiming to address the autonomous and personalized learning needs of students in the digital and AI era, as well as the requirements of the new engineering construction for the cultivation of advanced abilities in university physics general courses. By constructing a intelligent course teaching system and implementing teaching strategies such as problem oriented, project-based, and seminar based, combined with course specific intelligent agents to optimize students learning experience, and using big data to analyze students learning trajectories, personalized learning paths are recommended. Teaching practice has shown that the construction of smart courses has significantly improved students learning effectiveness, the overall improvement in mid-term and final grades in the past five teaching semesters is significant. The research conclusion shows that the construction and implementation of intelligent courses based on knowledge map and AI technology can effectively improve teaching efficiency and effectiveness, providing useful practical references for higher education reform in the new era.

Key words: intelligent course, knowledge mapping, blended learning, multi-assessment