AI big model and text mining-driven framework for urban greening policy analysis.

Policy analysis is essential to improving the rationality and adaptability of policies. Traditional policy analysis easily generates biased results due to different individual perspectives and personal experiences. Text mining emerges as an efficient way, but is not widely used in urban greening policy analysis. Moreover, existing policy studies are mostly limited to topic categorization, and there is a lack of systematic policy text analysis and real-time policy tracking. Here, we constructed a multidimensional dynamic policy analysis framework for systematic evaluation of urban greening policies by introducing AI big models and text mining. With Wuhan as an example, the framework was used to analyze the evolution of policy topics, distribution of annual topics, and spatial and temporal changes in greening indicators. Moreover, the framework supports real-time tracking and in-depth interpretation of policies, and the results can be presented through visualization scenarios. Analysis with the framework revealed variations of greening policies in Wuhan over the past 15 years, such as transformation from basic greening to ecological remediation and policy focus shift from flower planning to wetland protection. This methodology marks a new paradigm for intelligent policy evaluation, and significantly improves the efficiency and accuracy of policy formulation and implementation in smart cities.