Adaptive Learning Path Optimization Based on Reinforcement Learning in Chinese Teaching and Learning

Abstract

 In order to improve the teaching effect of Chinese courses, this paper focuses on the construction and assessment of adaptive learning paths, aiming to realize personalized teaching and meet students' diversified learning needs by using artificial intelligence technology. The article establishes a learner model from the learner's individualized needs, and designs a domain knowledge structure model in combination with the knowledge characteristics of the Chinese curriculum. The DINA model is used to diagnose students' mastery of Chinese knowledge points, and then a gated loop unit is introduced to design a knowledge tracking model for exploring students' knowledge level. Combined with the Q-Learning algorithm of reinforcement learning, the adaptive learning path recommendation method for Chinese knowledge was designed, and the data analysis was carried out through application practice. The results show that the performance of students' Chinese knowledge tracking level and path recommendation is better than the current mainstream methods, and the overall language learning performance of the experimental class students in the teaching practice has been improved by 14.69 points, and the mean value of the learning path score is 6.47 points (7-point scale). Therefore, relying on intelligent algorithms to assist the teaching of language courses helps to improve the level of language teaching and also provides guidance for innovative language teaching methods.