地方政府人工智能产业政策文本量化研究

通过构建政策"目标-工具"分析框架,借助语义本挖掘及社会网络分析等方法对检索得到的省级政府人工智能政策文本进行量化分析。得出结论:(1)人工智能政策的主题词呈现集中化和群聚化的特点;(2)各省政策目标主要围绕"发展环境""融合应用"和"产业化",但不同省份之间存在一定的差异性;(3)环境面工具应用过多,需求面工具使用过少,部分细分政策工具使用过于频繁;(4)人工智能政策中,政策目标和政策工具存在着某种程度的错配现象。最后,依据上述结论,提出政策建议。

A Quantitative Study on the Text of Local Government Artificial Intelligence Industry Policy

By establishing policy"target-tool"analysis framework,and using the method of text mining and social network analysis,the paper quantitatively analyzes the texts of provincial government artificial intelligence policies. The main conclusions are as follows:( 1) The keywords of artificial intelligence policy shows centralization and clustering characteristic;( 2) The policy objectives of each province mainly focus on " development environment"," integration application" and " industrialization",but there are certain differences among different provinces;( 3) There are too many applications of environmental surface tools,while the demand surface tools were used too little,and some segmentation policy tools are used too frequently;( 4) In the artificial intelligence policy,there are some mismatches between policy objectives and policy tools. Finally,based on the above conclusions,some specific policy recommendations were put forward.