Research on Data-Driven Student Personalized Learning Path Recommendation System

This study proposes a data-driven student personalized learning path recommendation system, which combines collaborative filtering algorithm with DNN. By analyzing students' learning behavior data, grades, interests and other multimodal information, a personalized student model is constructed. The system can dynamically recommend personalized learning paths based on students' learning history, learning preferences and behavior patterns. Through experiments on the learning data of 2,000 students, the system demonstrates its efficiency and accuracy in learning progress prediction and path recommendation. The experimental results show that the system achieves 94.2% in recommendation accuracy, which is 15% higher than traditional recommendation methods (such as content-based recommendations). In addition, the system's personalized recommendation of student learning paths also effectively improves students' academic performance. The scores of students using personalized paths are <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{1 8 \%}$</tex> higher than those of the traditional path group. These experimental results verify the significant advantages of the recommendation system in optimizing students' learning effects.