Optimizing regional disparities in green building evaluation: A random walk-linear regression analysis in China

China's vast geographic expanse encompasses significant regional variations in climate, economic development, and resource endowment. These disparities complicate the effective utilization of regional climatic and resource conditions for energy-saving efforts in green building evaluations under generalized systems. This study aims to construct a green building evaluation system that accounts for regional differences across China. Through a combination of literature review and questionnaire surveys, 14 regional characteristic influencing factors and 36 green building evaluation points were identified. The relationships between the influencing factors and evaluation points were established using hesitant fuzzy set theory and the analytical hierarchy process. A weight modification model based on the Random Walk-Linear Regression algorithm was then developed to quantify the impact of regional differences and dynamically adjust the weights of the green building evaluation system. By applying this model to adjust the evaluation indicator weights for China's provincial-level administrative regions, the results revealed differences in weight adjustments across regions, reflecting their distinct climate, economic, and resource conditions. Furthermore, dynamically adjusting the green building evaluation system to account for specific regional characteristics enhances energy efficiency, reduces costs, and promotes sustainability in green building practices. This study underscores the importance of tailoring green building evaluation systems to regional attributes and provides a theoretical foundation for policymakers to refine assessment systems and foster regionally adapted, sustainable green building development.