情報処理学会 インタラクション2026

This study demonstrates a data structuring methodology to expand the complex knowledge system of Wajima-nuri?a traditional Japanese craft?into a Virtual Reality (VR) learning space. Traditional craft knowledge is multi-layered, making it difficult to grasp interrelationships through linear text-based learning. While VR offers spatial immersion, effectively utilizing this medium requires restructuring unstructured data into information granularities capable of spatial arrangement. We utilized Generative AI agent to extract "Atomic Facts" from specialized documentation and constructed a Knowledge Graph based on a domain ontology. The results indicate that this structured data serves as a logical foundation for designing information nodes and visual guidance paths within VR, proposing a dynamic knowledge system that enhances information reception efficiency.