Using Geolocalized Social Network for Contact Tracing Based on Community Detection

Contact tracing is a method that traces infected people and then identifies all their contacts defined as potential new cases. In response to sudden outbreaks of epidemics, many contact tracing systems were proposed. Despite their efficiency, most of these systems are limited. While mobile applications require users' cooperation to install them, systems based on online social networks are more appropriate for communities active on online platforms. However, Google Maps with over twenty million downloads, has recently evolved into a social network in disguise where users share their opinions about the places they have visited. Thus, Google Maps data offer a good substitute for conventional contact tracking data. In this paper, we propose a novel approach applying a community detection algorithm to a social network built on Google Maps data. The proposed approach uses the Louvain algorithm to group both infected and safe people in communities, then identifies suspected cases based on resulted communities. Finally, the implementation of our approach shows its efficiency in covering more suspected cases compared to similar work.