Uncovering key topics and causal structures in cultural heritage governance using topic modelling and the DEMATEL approach

Effective governance of cultural heritage has become an increasingly complex challenge, involving diverse actors, overlapping jurisdictions, and competing values. Addressing these challenges requires a deeper understanding of the thematic structure and causal mechanisms that underpin heritage policy and management. This study employs natural language processing techniques to extract and analyse governance-related topics in the cultural heritage domain. Using Latent Dirichlet Allocation (LDA), six key topics are identified: guidance of cultural activities, legalisation of cultural heritage value, presentation of cultural heritage, multi-level governance of cultural heritage, activation of heritage space, and application of heritage protection technology. To explore the interrelationships among these topics, the Decision-Making Trial and Evaluation Laboratory (DEMATEL) method is applied to uncover the causal links and structural dependencies. The findings reveal a complex, multi-layered governance landscape that connects normative frameworks, participatory practices, spatial strategies, and technological interventions. This integrated analytical model not only highlights critical leverage points for improving heritage governance but also provides a replicable method for systematically identifying thematic priorities and influences. The research makes theoretical contributions to cultural heritage studies and provides practical insights for policymakers and practitioners seeking to enhance the coordination, effectiveness, and sustainability of heritage management systems.