In recent years, both in Japan and abroad, the digital archiving of rare materials and books has progressed rapidly. However, in current digital archive systems, achieving these goals faces two major challenges.
(1) Because expert judgment is still required to make academic resources searchable on a unified platform, the process of assigning metadata to digitized materials suffers from the same limitations as annotation in machine learning—namely, a shortage of human resources.
(2) For modern and contemporary materials, libraries, archives, and museums operate under different principles of public access, making it difficult to establish unified cross‑disciplinary usage policies.
To address these issues, this study will develop methods for structuring digital archives as data that can be utilized by generative AI, within an international collaborative research framework. Building on this foundation, the project will construct an international model for knowledge infrastructure in digital scholarly environments and articulate a future direction for the humanities in Japan in the DX‑AI era.
