Community Detection Using Graph Partitioning

Detecting communities in big networks is a tough issue. Therefore the community detection algorithms help us to partition the network into different communities. Community detection has been helpful in a wide range of domains. Particularly in the health domain, community detection can be used for finding the dynamics of certain groups or communities which are vulnerable to an epidemic disease. COVID-19 is a disastrous pandemic that swept across the globe, causing major problems in the world. It’s far swiftly spreading throughout the world, after its preliminary outbreak in China. The research communities around the world are carrying out various researches and studies to analyse the impact of COVID-19 on medical, economical and sociological fields. This paper introduces an enhanced Girvan-Newman algorithm where multiple-edge removal is possible. The improved algorithm is compared with the existing one and in practice, it reduces the number of iterations for each edge removal. Therefore the improved algorithm gives a faster community detection. The most important feature of our paper is that we are able to find different regions or communities which are affected by COVID-19 using the enhanced Girvan-Newman algorithm. Also, Visual evaluation of plots and maps are applied to present a geographical illustration of COVID affected communities.