Analysis of Foot Traffic due to COVID-19 using Mobile Network Big Data

Abstract

The COVID-19 pandemic has hit the tourism industry with a drop in the number of domestic and foreign tourists and loss of revenue. We can prepare for future pandemics by analyzing the impact of COVID-19 on tourist numbers before and after the outbreak. In this paper, we analyze the change in the foot traffic of tourists and resident population before and after the outbreak of COVID-19 using mobile communication company call records (CDR) and WiFi access log data. Analysis of changes in the foot traffic due to COVID-19 is analyzed based on the number of confirmed cases, recoveries, and deaths by administrative dong, daily, and gender. The hierarchical cluster analysis of distribution population for administrative dongs (classified as k = 5 clusters) was performed to classify areas with high changes in foot traffic due to COVID-19. We also predict the change in the foot traffic due to the epidemic using the Xgboost (Extreme Gradient Boosting) algorithm.