Identification and analysis of urban influential regions using spatial interaction networks

Understanding and controlling spreading processes in urban space is of theoretical and practical importance for developing policies in sustainable urban management. Conventional studies are mainly focused on identifying influential nodes in different networks, but little attention has been paid to identification of influential regions and analysis of their spatiotemporal characteristics in urban space. To fill this gap, this study proposes methods to identify urban influential regions and explores their spatiotemporal patterns using networks built from massive human movement data in Shanghai, China. (a) Influential regions in different time periods are identified as statistically significant hotspots of influential nodes, where betweenness centrality is used to extract the minimal set of influential nodes. (b) Influential regions are broadly distributed in urban space and exhibit periodic morphological patterns. Additionally, (c) they are associated with different urban functionalities and show varying degrees of influence in space, given their strong connections with the remainder of urban space either locally or globally. Lastly, (d) we report that they have different degrees of temporal influence with stable daily variation. Our results suggest the resilience and temporal variation of influential regions, and how this can be valuable in urban management such as disease control or resource circulation.

Automated summary

Using massive human movement data, this study proposes methods to identify urban influential regions and finds that they are broadly distributed in urban space, exhibiting periodic morphological patterns. These influential regions are associated with different urban functionalities and have varying degrees of influence in space, while also showing stable daily variation.