AVAIL: An AI-Driven Vocabulary Acquisition and Instructional Learning Framework

Abstract

The rapid advancement of generative artificial intelligence has unlocked new possibilities for language learning, particularly through intelligent agents powered by large language models (LLMs). This study explores the integration of the Coze agent with the ARCS motivational design model in the context of English vocabulary instruction. The research begins with a review of the ARCS model’s relevance to language education and the latest developments in AI-driven vocabulary acquisition, emphasizing the value of combining motivational frameworks with intelligent agents. A vocabulary learning pathway was created based on the four motivational dimensions of the ARCS model-Attention, Relevance, Confidence, and Satisfaction. By utilizing the Coze agent’s ability to create authentic learning environments and provide personalized feedback, the study designed a closed-loop, adaptive instructional approach. A six-week experimental study conducted at T Middle School found that students using the agent-supported learning pathway showed significant improvements in vocabulary performance and motivation compared to those using traditional methods. The results confirm the potential of LLM-driven intelligent agents in fostering sustained motivation, enhancing vocabulary transfer, and enriching the overall learning experience. This study proposes a comprehensive framework for the future use of intelligent technologies in language education.