Analysis and Recommendations for the Transformation of Korean Digital Cultural Heritage Archives into AI Training

This study analyzes the current status of Korea’s representative digital archive, e-Museum. Also, the proposes specific strategies for its development into an AI-enabled digital archive. At present, e-Museum remains dependent on a hierarchical classification system centered on function and material, a keyword-based search mechanism, and inconsistent metadata. As a result, artifacts are reduced to simplified objects devoid of their historical and social contexts, which is limiting both AI training and international interoperability. To overcome these limitations, the study employed four research methods. First, international guidelines issued by UNESCO, ICOM, and ISO, as well as key prior studies on AI and digital archives. Second, comparative analyses were conducted between e-Museum and major international platforms such as the Smithsonian, Europeana, and the F itzwilliam Museum. Third a pilot test was conducted by applying semantic prototype to selected artifacts. The results indicate that the development of e-Museum into an AI-based digital archive requires four key directions: ensuring data accuracy and consistency, strengthening the integration of visual information, securing interoperability through international standards, and establishing interdisciplinary design strategies. In addition, effective AI application depends on four concrete improvements: the incorporation of multi-layered classification structures, the adoption of semantic search and personalized recommendation algorithms, the construction of training datasets based on visual resources, and the establishment of expert validation mechanisms. These findings demonstrate that the proposed strategies are not abstract suggestions. Actionable directions, however, are grounded in international case studies and empirical analysis. In conclusion, this study emphasizes that e-Museum must evolve into a context-centered knowledge ecosystem that ensures both scholarly reliability and public usability in the AI era. Such a transformation will not only enhance the academic and practical value of digital cultural heritage in Korea but also contribute to international digital archives.