Population mobility,urban centrality and subnetworks in China revealed by social sensing big data

Abstract

Studying the spatial pattern of population flows is important to gain insights into economic connections, city networks and traffic demand. The dearth of integrative data suggests that the spatial pattern of population migration has still received relatively scant research with modified algorithms despite its vital guidance in epidemic control. To address this gap, we employ unique social sensing data from the Tencent migration platform to investigate the spatial configuration of population migration in China. Based on the identified 2,555,596 aggregated origin-destination records in 2018, we map the spatial pattern of population flow with the algorithm of eigenvector centrality. The geovisualisation illustrates that population flows present an uneven spatial pattern with a clear east/west divide. The geovisualisation depicts that the cities with higher administrative levels show corresponding high centrality to a certain extent. The results also show that subnetworks of migration flow demonstrate that population flows are near movements.