ABSTRACT
It is imperative to understand human movement and behavior, from epidemic monitoring to complex communications. So far, most research and studies on investigating and interpreting human movements have traditionally depended on private and accumulated data such as mobile records. In this work, social network data is suggested as a proxy for human mobility, as it relies on a large amount of publicly accessible data. A mechanism for urban mobility mining and extraction scheme is proposed in this research to shed light on the importance and benefits of the publicly available social network data. Given the potential value of the Big Data obtained from social network platforms, we sought to demonstrate the process of analyzing and understanding human mobility patterns and activity behavior in urban areas through the social network data. Human mobility is far from spontaneous, follows well-defined statistical patterns. This research provides evidence of spatial and temporal regularity in human mobility patterns by examining daily individual trajectories of users covering an average time span of three years (2018 to 2020). Despite the diversity of individual movements history, we concluded that humans follow simple, reproducible patterns. Additionally, we studied and evaluated the effect of COVID-19 on human mobility and activity behavior in urban areas and established a strong association between human mobility and COVID-19 spread. Numerous years of mobility data analysis can reveal well-established trends, such as social or cultural activities, which serve as a baseline for detecting anomalies and changes in human mobility and activity behavior.
ABSTRACT
Over the past few months, COVID-19 has emerged to the world as a new threat to humanity and communities, expanding from a few small infected cities to hundreds of countries around the world impacting businesses, education, economics, and almost every activity associated with human life. This had led many researchers and scientists to analyze and study different factors and variables that obtain timely information on the outbreak of COVID-19. One of the main factors that helped in spreading the corona-virus is human mobility. Since detailed information about human movement during outbreaks are difficult to obtain, social networks comes as an alternative with its massive volume of publicly available data. In this research, we propose mobility detection and identification of social media's spatio-temporal data, as a proxy for human mobility. We aim to discover and explore an in-depth level of mobility data extracted from social media applications to uncover the relation between COVID-19 spread and daily mobility ratio in Kuwait regional area. With the use of the latest mobility data extracted from Twitter users, we have shown that user mobility is linked to the positive cases of COVID-19, with a relatively high correlation coefficient. Moreover, we have analyzed and discussed how the impact of COVID-19 affected user behavior and mobility habits. © 2020 ACM.