Robustness analysis of the two-mode COVID-19 network

Abstract The weighted complex networks have been extensively used to model real-world phenomena including epidemic diseases, to analyze the relationships within their units namely nodes and links. COVID-19 is one of those infectious diseases which demands serious attention to mitigate its diffusion. According to the facts and figures by the WHO, more than 775 million people are at life-threatening risk because of COVID-19. Various research studies have been suggested to cope with the COVID-19 pandemic. However, this study aims to understand the COVID-19 cases diffusion problem as a network, examining its robustness and vulnerability to random attacks. For this, we removed links between the nodes in this two-mode network using a random and targeted approach by applying different centrality measures (degree, closeness, betweenness, and eigenvector). Additionally, these centrality measures helped identify this network’s focal (hub) nodes. The results show that merely 15% of targeted attacks are equal to 76% of random attacks for this two-mode network. Thus, the robustness of nodes and links purely depends on the topological aspect of the network. Finally, the random attacks have a negligible impact on the robustness of the COVID-19 network; however, the targeted attack is productive in removing nodes with a higher degree.