我国人工智能政策特征与PMC模型量化评价研究

Research on policy characteristics of artificial intelligence and PMC model evaluation in China

摘要:

探讨我国人工智能政策特征以优化政策设计思路，以中央部委与地方政府印发的31份人工智能政策文件为研究样本，综合运用文本挖掘、内容分析和PMC指数模型等方法，提炼人工智能领域的研究前沿与热点分布、人工智能的政策特征以及评价结果。研究发现：我国人工智能政策以智能化为主要目标，重点关注科技研发与创新环节，并对“卡脖子”的芯片、机器人产品提出政策规范；政策工具应用不平衡，对政府采购、对外贸易管制、反他国管制、购置补贴四项需求型政策工具重视度不够；总体政策效力较好，但政策之间具有效力差异性，相关政策条款亟待加强和完善。基于以上特征和问题，建议进一步促进人工智能前沿政策的设计与前瞻，加大需求端政策工具的组合性和灵活性应用、优化供需政策工具包，加强对人工智能领域的监管政策力度，开展人工智能政策系统的整体性评估。

Abstract:

This paper discusses the characteristics of AI policy in China, takes 31 AI policy documents issued by central ministries and local governments as research samples, and comprehensively uses the methods of text mining, content analysis and PMC index model to refine the research frontier and hot spot distribution in the field of AI, the policy characteristics and evaluation results of AI. It is found that China's AI policy takes intellectualization as the main goal, focuses on scientific and technological R & D and innovation, and puts forward policy norms for "neck stuck" chips and robot products; The application of policy tools is unbalanced, and insufficient attention is paid to the four demand-based policy tools: government procurement, foreign trade control, anti other country control and purchase subsidy; The overall policy effectiveness is good, but there are differences in effectiveness between policies, and the relevant policy provisions need to be strengthened and improved. Based on the above characteristics and problems, it is suggested to further promote the design and foresight of AI cutting-edge policies, increase the combination and flexible application of demand side policy tools, optimize the supply and demand policy toolkit, strengthen the supervision and policy in the field of AI, and carry out the overall evaluation of AI policy system.