Development of a Generative AI-Based Platform for Construction Education by Integrating Retrieval-Augmented Generation

The digitalization of the construction industry brings new opportunities for improving productivity, coordination, and information exchange. However, this digital transformation has reshaped construction management knowledge into a more interdisciplinary and fragmented form, making it harder for educators to structure coherent, standards-aligned instruction. Recent advancements in generative artificial intelligence (AI) offer potential solutions by supporting adaptive, context-aware, and scalable learning environments. Grounded in constructivist learning theory (CLT), this study proposes and examines a pedagogical framework for construction management education that integrates retrieval-augmented generation (RAG) to enable verifiable, context-aware learning interactions. Unlike general-purpose AI tools, the framework confines retrieval to course materials and structures responses around reference-grounded guidance. Evaluation through expert review, automated scoring, and a formative student study produced evidence of educational and functional effectiveness, expert ratings averaged 4.5/5, automated metrics averaged 0.96/1, and the student study (n=7) confirmed usability and instructional value across relevance, accuracy, clarity, and scope alignment. These findings demonstrate that the framework supports transparent, evidence-based learning. Built on a constructivist foundation, it demonstrates how generative AI can be pedagogically applied to strengthen teaching and learning in construction management and advance the broader goals of engineering education and development.