Research on the PMC Index Evaluation of China's Artificial Intelligence Policy Based on "Policy Tools + LDA": An Empirical Study from China

Abstract

As a transformative technology leading the future, the rapid development of Artificial Intelligence (AI) brings both enormous opportunities and governance challenges, making rational policies crucial for its healthy growth. Strengthening analysis and evaluation of policy texts provides a basis for policy optimization. Existing studies mainly focus on policy evolution, efficiency, and topic mining, lacking in-depth analysis of the internal characteristics of policy tools and quantitative assessment of policy merits and demerits. Methodologically, they seldom adopt quantitative approaches like policy index analysis, and traditional methods struggle to balance analytical depth with evaluation scientificity. Therefore, this study constructs a comprehensive PMC index evaluation model integrating "policy tools + LDA topic model."Taking 213 central-local AI policy texts in China as the research object, it codes and classifies policy tools, conducts quantitative content analysis via LDA topic modeling, and selects 10 representative policies for PMC index empirical evaluation. Based on the findings, it proposes recommendations to optimize policy tool structures, improve industry regulatory systems, and strengthen policy implementation safeguards. The research aims to upgrade China's AI policy framework, promote high-quality industrial development, and facilitate standardized applications.