Characterizing the spatial-functional network of regional industrial agglomerations: A data-driven case study in China's greater bay area

Abstract

Agglomeration economies have been an increasingly prevalent interest in research communities of regional and urban studies. Recent evidence reports the importance of agglomeration economies closely associated with the urban network. However, limited efforts have been proposed to emphasize the spatial-functional roles of regional agglomeration and their network behaviors. With the use of geospatial big data and network analytical approaches, this paper, therefore, presents a data-driven intraregional case study in China's Greater Bay Area (GBA) to investigate the curiosities related to a regional network of sectoral agglomeration economies. By extracting industry-related points of interest (POIs), industrial activities are derived by topic modeling in representing industrial functional topics. Bipartite network projection is implemented to construct a one-mode network based on the two-mode agglomeration network. Community detection is further applied over the agglomeration networks, for uncovering groups of agglomerations that are closely connected by their industrial functions. The results confirm substantial geographic heterogeneities of agglomeration networks related to locational advantages. A robust radiation effect on the north-western hinterland is observed regarding material and mechanical production industries. It is suggested that complementary relationships in regional cooperative activities have been partially established through intensive agglomeration networks.