The Strategy of Accurately Pushing Course Resources Using Artificial Intelligence Algorithms in Online Learning Systems

With the rapid development of online education platforms, users are faced with the problem of overloaded course resources and mismatched personalized needs, resulting in low learning efficiency and poor learning experience. This study first collects user behavior data and course metadata through crawler technology, and performs data cleaning and normalization; secondly, it constructs user portraits and course feature vectors, including learning preferences, knowledge level, course difficulty and other dimensions; then, it uses collaborative filtering algorithms and deep neural network models to train personalized recommendation models, and combines reinforcement learning to dynamically adjust push strategies. The test finds that the precise push strategy based on artificial intelligence algorithm significantly improved the learning effect of users, the course completion rate increases to 100%, and the daily learning time is as low as 1.6 h. This study confirms the effectiveness of artificial intelligence algorithm in accurately pushing course resources in online learning systems, and provides theoretical basis and practical guidance for the personalized service optimization of online education platforms.