本研究基于技术挖掘方法，依托DerwentInnovation专利数据库，引入以专利申请数量和专利授权数量为代表的数量型指标、以向外布局专利数量和专利被引情况为代表的质量型指标和以联合申请和技术相似为代表的网络化指标三个维度，剖析全球人工智能的技术布局和竞争态势，进而探究34个代表性城市（都市圈）的创新能力。研究发现：一是在人工智能领域中国城市（都市圈）在数量型指标方面位于全球前列，并且这种规模优势呈现愈加明显的态势；二是中国城市（都市圈）人工智能的技术创新质量与欧美国家相比存在较大的差距，尤其是向外布局专利数量增长趋势开始出现“逆势下降”的状态；三是中国城市（都市圈）之间维持着较强且稳定的技术合作强度，国际技术合作强度远落后于美国和日本，合作网络位势的优势并不十分突出。
Based on global patent data from the Derwent Innovation patent database，this study introduces Tech Mining to explore the innovation capacity of global representative urban agglomeration on Artificial Intelligence（AI）by introducing three dimensions：quantity indicators represented by the number of patent applications and patent grants，quality indicators represented by the number of oversea patent applications and overall patent citations，and networking indicators represented by joint applications and technology similarities. The study also explores the innovation capability of 34 representative urban agglomerations. It is found that firstly，the urban agglomerations in China are at the forefront of the global AI field in terms of quantitative indicators，and this scale advantage is becoming more and more obvious；Secondly，there is a big gap between the quality of AI technology innovation in Chinese urban agglomerations that in Europe and the United States，especially in the number of oversea patents，which has started to show a “counter-trend decline” trend；Thirdly，China’s urban agglomerations maintain a strong and stable technological cooperation，while the intensity of international cooperation lags far behind the United States and Japan，and the advantages of the cooperation network are not very prominent.
