Intelligent Educational Recommender System Using Deep Learning and Natural Language Processing for Personalized Learning Paths

This research presents an intelligent educational recommender system that leverages deep learning and natural language processing techniques to create personalized learning paths for students. The proposed system addresses the challenge of individualized education in digital learning environments by analyzing student behavior patterns, learning preferences, and knowledge gaps. We developed a hybrid neural network architecture combining Collaborative Filtering with Content-Based Filtering using LSTM and BERT models to process educational content and student interactions. The system was evaluated using the EdNet dataset containing ۱۳۱ million educational interactions from ۷۸۴,۳۰۹ students. Our experimental results demonstrate that the proposed system achieves ۹۲.۴% accuracy in predicting student performance and ۸۹.۷% precision in recommending appropriate learning materials. The system successfully reduced average learning time by ۲۳.۵% while improving knowledge retention rates by ۱۸.۲% compared to traditional learning approaches. The integration of natural language processing enables the system to understand textual educational content and match it with student learning styles effectively. This research contributes to the advancement of intelligent tutoring systems and provides a scalable solution for personalized education in both traditional and online learning environments.

کلیدواژه ها:

Educational Recommender Systems ، Deep Learning ، Natural Language Processing ، Personalized Learning ، LSTM ، BERT