Abstract: In order to have a deeper grasp of the content of artificial intelligence (AI) policies at the central and local levels, and clarify the characteristics of AI policy layout in different regions, the paper mines and quantitatively analyzes the content of 11 AI policy texts at the central level, and constructs a three-dimensional analysis framework of “subject(Y)-goal(Z)-tool(X) ”of China’s AI policy. Under this framework, the content of 84 AI policy texts at the local level was mined and quantitatively analyzed, and the differences in the association of “Y-Z”“Y-X” and “Z-X ” of AI policies in the eastern, central, western and northeastern regions of China were compared. The results show that in terms of “Y-Z” correlation, the government, enterprises, universities and scientific research institutions are all involved in the realization of various policy goals, but the participation of the government is relatively limited. In terms of “Y-X” relevance, the correlation structure between enterprises and policy instruments in each region is similar, and all of them are mainly supply-oriented tools, taking into account environment-oriented tools. The correlation between the government and policy tools is generally low, and the role of local governments has gradually evolved from a leader to a guide. The relationship structure between universities and scientific research institutions and policy instruments in different regions is similar, and all of them are mainly supply-oriented tools, taking into account demand-oriented tools. In terms of “Z-X” relevance, the combination of policy tools used in each region is relatively simple, and supply-oriented tools are used to achieve “technology leadership” and promote “industrial transformation”. the use of demand-based tools to achieve “theoretical breakthroughs”; We use environment-based tools to ensure “ethical safety”.

Key words: Artificial Intelligence Policy; Text Mining; Text Relevance; Policy Participants; Policy Objectives; Policy Tool