Machine Learning-Driven Topic Modeling and Network Analysis to Uncover Shared Knowledge Networks for Sustainable Korea–Japan Intangible Cultural Heritage Cooperation

In this study, we provide a comparative analysis of intangible cultural heritage (ICH) research trends in Korea and Japan, aiming to uncover shared knowledge networks and potential areas for sustainable cooperation. We employ a mixed-method approach, combining machine learning-driven topic modeling using Latent Dirichlet Allocation (LDA) and network analysis techniques, to examine a corpus of Korean and Japanese research papers on ICH. LDA topic modeling identified three primary themes: technology and ICH, safeguarding ICH, and methodologies and approaches in ICH research. Comparative analysis reveals distinct characteristics in each country’s approach. Korean research emphasizes practical applications of technology and policy-driven safeguarding strategies, while Japanese research leans towards theoretical exploration and cross-cultural comparisons. Citation network analysis further identifies influential papers and shared knowledge bases, underlining potential opportunities for collaboration. Key findings highlight the potential of technology for ICH preservation and promotion, the necessity of comprehensive safeguarding strategies, and the crucial role of community engagement. Our study suggests that by leveraging their complementary strengths and engaging in collaborative research, Korea and Japan can contribute to the sustainable safeguarding of ICH and foster a deeper understanding of their shared cultural heritage.