A multidimensional heritage value assessment framework for ancient Chinese medical texts: evidence-based prioritization of preservation and utilization

Purpose This study develops an objective, quantitative, and technology-driven multidimensional framework for assessing the heritage value of ancient Chinese medical texts. The framework aims to overcome the limitations of traditional expert-based approaches and support evidence-based conservation and sustainable reuse. Design/methodology/approach The proposed framework integrates multi-source data analysis with large language models (LLMs) and latent Dirichlet allocation (LDA) to quantify scholarly value, contemporary practical value, historical value and cultural influence. Indicator weights are determined using a hybrid weighting approach that combines expert judgment and objective data. In addition, an artistic value feature identifier and a physical-condition matrix are incorporated to support conservation and utilization decisions. Findings An empirical assessment of 200 ancient Chinese medical texts shows that scholarly value and contemporary practical value contribute more substantially to overall heritage value than traditional attributes such as age. The results also reveal a highly right-skewed value distribution, with 60% of the sampled texts classified as Level IV. The framework effectively identifies high-priority texts, such as analytic comparison of cold damage and warm pestilence, for targeted conservation and management, thereby supporting dynamic resource allocation. Originality/value By introducing a reproducible, multidimensional framework, this study shifts the assessment of ancient texts from manual judgment to an evidence-based, technology-driven approach. The framework offers a scalable and adaptable methodology for the intelligent management, conservation and utilization of documentary heritage.