Currently, personalized learning supported by big data has become a research hotspot in the field of education. This study applies data mining technology to English learning and constructs a personalized path generation system for English learning by using online learning data from new media teaching environment. Through learner profile modeling and personalized learning resources recommendation, we generate learning paths adapted to students’ level and learning habits. The high accuracy of the personalized learning resource recommendation method in this paper is verified through experiments, and its check accuracy and check completeness rates are 6.35%~22.23% and 12.91%~27.35% higher than those of the comparison methods. Students’ pass rate of the English general examination is increased to more than 70% after applying the system, and most of the students agree in the personalization, learning effectiveness dimension and behavioral willingness dimension, which reflects students’ good satisfaction with the system. The study shows that the personalized path generation system for English learning based on data mining in this paper is able to stimulate students’ interest and effectively improve their learning motivation and English learning level.
