Entropy-based evaluation of architectural components in traditional Huizhou dwellings: a data-driven approach to heritage conservation

Architectural heritage conservation increasingly requires objective and reproducible approaches to evaluate the cultural and physical significance of traditional dwellings. Despite the growing emphasis on authenticity and integrity in global heritage frameworks, quantitative tools for assessing these values remain limited. To address this gap, this study introduces a data-driven evaluation model employing the Entropy Weight Method to evaluate the heritage significance of architectural components in traditional Huizhou dwellings. Field surveys were conducted on 35 traditional dwellings across six representative villages in the Huizhou region, focusing on 17 categories of architectural components. The methodology comprises four key steps: standardized scoring of preservation status and cultural value, data normalization, entropy-based weighting, and hierarchical classification. Results reveal significant variation in both the physical condition and cultural symbolism of the assessed architectural components. High-ranking components, including horse-head gables, patios, and Huizhou-style carvings, exhibited the highest scores and weights, highlighting their roles in both physical conservation and cultural expression. Conversely, components with limited visual prominence and low variation, such as drum-shaped stones and beauty rests, received notably lower rankings. The proposed framework enhances transparency and comparability in heritage value assessment, providing a transferable reference for tiered conservation planning and value-based management across different regional contexts.