Research on the Application of Intelligent Recommendation System in English Reading Teaching

This paper explores the application of an intelligent recommendation system based on the Transformer model in English reading instruction to enhance students' personalized learning experiences. Through offline experiments with the reading data of 5,000 students and online experiments involving 300 students, the results indicate that the Transformer model, with an accuracy of 0.82, a click-through rate of 0.45, and an NDCG of 0.85, significantly outperforms traditional recommendation algorithms. In the online experiments, the average reading time of students increased to 144 minutes, the click-through rate rose to 0.35, and satisfaction scores improved to 4.3. The findings demonstrate that this recommendation system is highly effective and feasible in practical applications.