現行教育體系普遍採用統一的教材與教學方式，忽略了學習者的個別差異，導致學習成效不佳。本研究旨在突破傳統的教育模式，來建構一套基於知識圖譜的Python程式設計精準教育系統。研究採用模組化整合訓練(MIAT)方法論，結合IDEF0與Grafcet設計系統架構，整合提示工程、知識圖譜與檢索增強生成(RAG)技術，建立個人化學習系統。實驗結果顯示，本系統通過分級教學機制成功將學習內容分為基礎級、中等級與進階級，使不同程度的學習者能獲得適合的學習內容；透過InstaGraph.ai視覺化分析，系統能根據學習者測驗結果動態建立個人化知識圖譜，有效識別知識盲點與認知偏差；與傳統學習方式相比，結合知識圖譜與RAG技術的個人化學習系統提升了學習效率，使新手學習者在短時間內能掌握基礎Python程式設計概念。本研究的主要貢獻在於將AI技術有效整合至精準教育系統中，提出一套可自動根據學習者表現動態調整的個人化學習架構，不僅為Python程式設計教育提供新思路，也為其他領域的精準教育模式提供了可推廣的實作範例。;The current education system commonly adopts standardized teaching materials and methods, often overlooking individual differences among learners, which leads to suboptimal learning outcomes. This study aims to break through traditional educational paradigms by developing a precision education system for Python programming based on knowledge graphs. The system is designed using the Modular Integrated Architecture Training (MIAT) methodology, incorporating IDEF0 and Grafcet for system architecture, and integrating prompt engineering, knowledge graph technology, and Retrieval-Augmented Generation (RAG). The personalized learning system constructed enables adaptive instruction tailored to individual learners.
Experimental results show that the system successfully classifies learning content into basic, intermediate, and advanced levels through a tiered instruction mechanism, allowing learners at different proficiency levels to access appropriate content. Using InstaGraph.ai for visual analysis, the system dynamically constructs personalized knowledge graphs based on learners′ test results, effectively identifying knowledge gaps and cognitive biases. Compared to traditional learning methods, the integration of knowledge graphs and RAG technology significantly improves learning efficiency, enabling novice learners to grasp fundamental Python programming concepts in a shorter time.
The main contribution of this research lies in the effective integration of AI technologies into a precision education system. It proposes a dynamic, performance-based personalized learning framework that not only offers a novel approach to Python programming education but also provides a scalable model for precision learning in other domains.
