全球城市知识流动网络的结构特征与影响因素

在科技全球化时代,城市之间的知识流动日益频繁,成为当代知识生产的重要特征.然而,鲜有从知识流动的视角开展全球城市体系的研究.基于2017年的高被引论文合作数据,采用社会网络分析方法和空间计量模型系统地刻画了全球城际科研合作网络的拓扑结构和空间格局及其影响因素.研究发现:①拓扑结构呈现出以北美、欧洲和亚太城市的三极格局,...

Structural characteristics and influencing factors of the global inter-city knowledge flows network

In the age of globalizing science and technology, urban economic development increasingly rests on knowledge production and knowledge flows. Inter-city scientific collaborations, as the most potent aspect of modern knowledge production, are more and more frequent, which produces some of the highest quality science. However, there is a paucity of analysis of the world city system from the knowledge flows perspective. Using highly cited papers data from the Web of Science database in 2017, this study applies social network analysis, a Bayesian-inference weighted stochastic block model (WSBM), the dominant flow analysis and spatial interaction model to explore the topological structure, spatial pattern and influencing factors of global inter-city scientific collaboration. Results show that the science hotspots are highly concentrated in three regions: North America, East Asia and Western Europe, and the whole network is dominated by a tri-polar world. The two seemingly paradoxical trends, star-shaped and triangulated structure, coexist in the global inter-city knowledge flows network. The spatial pattern of inter-city collaboration network forms a quadrilateral graph with four vertexes in Western Europe, North America, East Asia and Australia, particularly on the trans-Atlantic axis between North America and Western Europe. The network has a distinctive multicore-periphery structure, which can be divided into five categories: global core, macro-region core, strong semi-periphery, semi-periphery, and periphery, and identifies New York, London, Boston, San Francisco-San Jose, Washington, Los Angeles, Paris and Beijing as eight global core cities and forty-one macro-region cores. The network is characterized as hierarchical “hub-and-spoke” structures, and the hierarchy of the network is obvious, New York, Beijing and Jeddah are dominant nodes in the three subnetworks. In addition, the gravity model indicates the spatial distance impedes inter-city scientific collaboration, while the amount of science output and the number of urban residents, the number of world-class universities, institutional proximity and social proximity have positive and significant effect on inter-city scientific collaboration. In order to further our understanding of world city network, this paper calls for more attention to inter-city knowledge flows.