Research on Personalized Hybrid Recommendation System for English Word Learning

Our study aims to explore a personalized hybrid recommendation system for English word learning. This paper begins with an analysis of the research status and existing problems in the field of personalized English word recommendation, followed by the design of a personalized hybrid recommendation system for English words. The system considers factors such as the user's word mastery level, the forgetting curve, and examination frequency of words. It utilizes three intelligent technologies: word vectors, cosine similarity, and long short-term memory networks, providing an innovative solution for the problem of English word recommendation. The system's efficacy is validated through an offline simulation experiment, comparing its performance with four baseline algorithms. The results reveal that the system outperforms the baseline algorithms in multiple indicators, demonstrating its superior recommendation capabilities. These findings have significant application value for English word learning and contribute valuable insights to personalized English word recommendation.