Correlation analysis between Baidu migration network and COVID-19 epidemic in China

Abstract

Population mobility affected the spread and risk diffusion of COVID-19. Based on Baidu migration big data and COVID-19 cases data released by the national health commission of people's republic of China combined with mathematical statistics analysis and geographic information technology, OLS test and geographically weighted regression were used to analyze the correlation between the spread of COVID-19 and Baidu migration network from January 10 to March 14, 2020.The results showed that the diffusion process of COVID-19 epidemic in China was characterized by stages, including outbreak, potential diffusion, rapid diffusion, diffusion inhibition and diffusion reduction. In the study period, there is a certain spatial correlation between the COVID-19 epidemic data and the difference coefficient of inflow and outflow and the external connection degree of provinces. Through the OLS test of population migration index, it was found that the correlation between the difference coefficient of inflow and outflow and the spread of epidemic was more significant, and there was no collinear effect. The correlation analysis showed that there was a correlation between the epidemic data and the difference coefficient of inflow and outflow in spatial location, and most of them were negative correlation in the early stage, and gradually became positive correlation in the later stage. The negative correlation between Hubei and Hubei was significant, and the positive correlation between Xinjiang, Tibet and Qinghai was significant. It revealed that provinces with large population mobility and high number of confirmed cases were mainly distributed in Hubei Province and the central cities of China's key urban agglomerations, and the epidemic prevention pressure was mainly due to the risk of transmission and diffusion caused by large population mobility and high number of confirmed cases.