Application of Knowledge Graph Technology in the Construction of University Physics Knowledge System and Learning Path Recommendation

With the advancement of new engineering education, university physics has become a core foundational course for science and engineering students, but its knowledge system is highly abstract, cross-branch, and dynamically evolving. Traditional knowledge organization methods cannot adequately support intelligent teaching or personalized learning, and problems such as fragmented knowledge points, unclear conceptual connections, and weak learning-path guidance increase students’ cognitive load. To address these issues, this paper applies knowledge graph technology to construct a university physics knowledge system and develop personalized learning path recommendations. First, the core modules and knowledge associations of university physics, including mechanics, thermodynamics, and electromagnetics, are defined through ontology construction. Second, knowledge extraction, fusion, and graph visualization are implemented to clarify semantic dependencies such as the relation between Maxwell equations and electromagnetic wave propagation. Third, a personalized learning path recommendation algorithm is designed using student learning data and graph reasoning. A 16-week experiment involving 286 science and engineering students from three universities shows that the knowledge graph increases the knowledge-point mastery rate by 32.7%, while personalized path recommendation improves learning efficiency by 28.9% compared with traditional teaching.