A comparative analysis of Chinese green building policies from the central and local perspectives using LDA and SNA

China’s construction, energy consumption, and emissions are the world’s largest, making green building (GB) crucial for sustainable development and climate action. GB policies play a crucial role in promoting sustainable development and mitigating environmental impacts. This paper presents a comprehensive analysis of China’s GB policies at both central and local levels using natural language processing (NLP), Latent Dirichlet Allocation (LDA), and semantic network analysis (SNA). It also integrates three theoretical frameworks: Policy Implementation Theory, Sustainable Development Theory, and Innovation Diffusion Theory. The study aims to: (1) compare the characteristics of central and local GB policies; (2) identify problems and propose solutions to improve policy effects and impacts. The results show that both levels of policies cover various sustainability aspects but with different focuses. The SNA reveals the semantic relationships and differences within the policy documents. The study suggests several strategies to enhance GB policy effectiveness and sustainability, based on the three theoretical frameworks. These strategies include providing more specific guidance, enhancing financial support and incentives, promoting technological advancements, strengthening monitoring and evaluation, adopting a more systematic approach, democratizing policy formulation, identifying key stakeholders and influencers, and considering innovation characteristics. This study contributes to the GB policy research and practice in China. It is also the first study to use NLP, LDA, and SNA to comprehensively analyze China’s GB policies, and the first study to apply the three theoretical frameworks to GB policy analysis.