In the era of artificial intelligence, building an informatized educational management system will significantly enhance the level of educational management in universities. This paper outlines the construction ideas for an AI-driven educational management information system. The system incorporates technologies like association rules and the Apriori algorithm with different interest, which are applied to the relevance of university courses. It collects course data, analyzes association rules, and suggests learning suggestions. The paper also describes the application of the C4.5 decision tree algorithm in an academic early warning model and verifies through experiments that the achievement early warning model based on the C4.5 decision tree algorithm has an ROC area exceeding 0.85, which can effectively improve the accuracy of student academic early warnings. The system has the potential to assist students in learning well and improve the efficiency of school management. Constructing an automated educational management system driven by artificial intelligence contributes to promoting modernization of university management.