This research aims to solve some major problems in smart education systems. These include the difficulty of connecting related ideas that are far apart in meaning when creating knowledge graphs, and the low accuracy of recommending the right learning resources to students. To overcome these challenges, the study developed new tools that can automatically build and improve knowledge graphs, plan learning paths based on each student’s needs, and suggest helpful resources using advanced recommendation methods. These solutions help reduce confusion caused by too many scattered learning materials. They also make it easier to explore learning data and provide personalized guidance for students at a large scale. The diagram below shows how the system works.