An Empirical Study on the Effectiveness of an AI-Assisted Teaching Model for Chinese Constructional Grammar in Mobile Applications: Integrating Cognitive Load Theory in Second Language Acquisition

Abstract

Chinese construction grammar emphasizes the pairing of form and meaning and presents a teaching challenge in second language acquisition. This paper designs a teaching model for Chinese construction grammar based on the cognitive load theory of second language acquisition. First, a lightweight knowledge graph for Chinese construction grammar was constructed, defining four types of entities—construction names, concepts, examples, and structural components—along with their relationships, and mapping out the semantic network among constructions to reduce learners’ internal cognitive load. By embedding the knowledge graph into a learning platform and linking knowledge points, the model accurately identifies students’ knowledge gaps and recommends related concepts and practice exercises. By incorporating cognitive load theory in second language acquisition, we completed the construction of the Chinese construction grammar teaching model. Performance test results show that, compared to the best-performing KGAT model, the recommendation algorithm proposed in this paper achieved a 0.26% and 0.03% increase in recall, respectively, and a 0.18% and 0.35% increase in NDCG, respectively. Empirical experimental results show that the experimental group demonstrated significant improvements in both receptive and productive skills, while the control group showed a significant improvement only in receptive skills. This demonstrates that the application of the AI-assisted Chinese constructional grammar teaching model on mobile devices effectively reduces students’ cognitive load and enhances their learning efficacy in Chinese constructional grammar.