Research on the Optimization of Intelligent English Vocabulary Teaching Paths Based on Reinforcement Learning Models

Abstract Educational intelligence has entered a new stage of development and is shifting from digital education to intelligent education supported by modern technologies such as big data analysis and intelligent algorithms. The purpose of this paper is to propose an intelligent English vocabulary teaching path by combining the reinforcement learning method to personalize the learners’ English vocabulary learning process. The English vocabulary exercise recommendation assisted teaching system in the teaching path utilizes the decision framework of the reinforcement learning model and the Q-Learning algorithm, which can classify the learning ability level of the learners and evaluate the difficulty of different English vocabulary exercises, so as to personalize the recommendation of the exercises with a specific level of difficulty to the students with the same class of learning ability level. After experiments, the Precision@3 of the proposed algorithm in the Yelp dataset is increased by 20.60% compared with the better-performing algorithm. More than 80% of the students in the application survey of the system gave affirmation in the evaluation dimensions of satisfaction, learning attitude, learning effect, etc., agreeing that the system is helpful to their learning. It shows that the intelligent English vocabulary teaching path method based on reinforcement learning model proposed in this paper is scientific and provides new ideas for teaching path optimization.