Enhancing Python Learning Through Retrieval-Augmented Generation: A Theoretical and Applied Innovation in Generative AI Education

Hsing-Tzu Ko, Yen-Ku Liu, Yun-Cheng Tsai, Summit Suen

Abstract: Our study presents an innovative learning tool that leverages the synergy between Retrieval-Augmented Generation (RAG) and Large Language Models (LLMs) to redefine Python programming skill acquisition. This research explores how integrating RAG with LLMs like ChatGPT can overcome traditional learning barriers by providing precise, contextually relevant responses, streamlining learning, and boosting learner confidence. Implementing RAG-enhanced LLMs resulted in decreased cognitive load and enhanced grasp and application of complex programming concepts. Our findings suggest that this RAG-based tool improves information reliability and enriches learning experiences, fostering more profound understanding and robust confidence in tackling programming challenges. This study contributes to the discourse on AI-assisted learning by showcasing RAG’s potential to enhance programming learning efficacy and satisfy learners ethically and at scale.