Leveraging Social Network Analysis For Optimized Community Detection: Insights From Global COVID-19 Datasets

Abstract

This research paper explores social network analysis for detecting communities within COVID-19 dataset to enhance pandemic outbreak response and prediction. Advancing community detection techniques, especially in the early stages of the pandemic, enables an improved understanding of its emergence and trajectory, resulting in optimized response effectiveness. The study employs network modeling, data visualization, and correlation analysis to identify communities in global COVID-19 data.  Networks are created with case counts, deaths, and recoveries, and a community detection algorithm is applied to uncover clusters and pandemic trajectories. The clean and detailed COVID-19 data enables a comparative analysis of populations with varying mortality and recovery rates, enriching the community detection process. The paper also introduces a graph generator using Gaussian distributions to create realistic sparse networks, providing synthetic data for further analysis. The key contribution of this study is enhanced preparedness in identification, clustering, and tracking a potential pandemic. Future research will focus on dynamic networks to track the evolution of communities over time and integrate statistical models to enhance the realism of network structures.