COVID-19 Pandemic: Centrality & Modularity

Authors

  • Caballero, Brenda G., Paloma, Zenas B. , Sedon, Mariz Rochelle C.

Abstract

The 2019 novel (new) coronavirus has since spread globally, resulting in an ongoing pandemic. In light of this outbreak, the COVID-19 Inter-Agency Task Force (IATF) has placed the Philippines under Community Quarantine to manage the Corona Virus Disease situation. Before easing the implementation of community quarantine, the country needs a careful evaluation of the pandemic situation to prevent another wave of infection. Thus, the crucial question is, “which provinces have the highest risk of increasing COVID-19 cases?” This study sheds light on that question by analyzing the Network Measures of Centrality. Moreover, this study provides a visual representation of the influential provinces whose networks are crucial to the spread of COVID-19 infection. This study analyzed the transmission of COVID 19 in clusters of a network by generating a modularity structure, degree centrality, and eigenvector centrality. Modularity structure is the cluster of provinces who share a strong and dense connection between each other. The degree centrality is the province (edge) that have the highest risk of catching the COVID-19 infection due to its connection to provinces (nodes) who have high cases of COVID-19 infection. And the eigenvector is the province with high case of COVID-19 infection that are influential in transmitting COVID-19 to linked provinces. The Gephi software was used for network visualization and analysis. Results show that, high levels of modularity can be observed. There are 7 clusters generated from the graph. The first cluster accounts for 30.23% of the total number of provinces. The 2nd cluster accounts for 17.44%, the 3rd cluster 9.3. The Degree centrality shows that cluster 1 has the highest degree centrality wherein Cebu is linked to 32 provinces, NCR to 14 provinces, Laguna to 7 provinces, Davao to 5 provinces, Cagayan Valley to 4 provinces, Cotabato to 3 provinces, and Zamboanga linked to 2 other provinces. Moreover, cluster 1 is also the eigenvector which consists of influenctial nodes that can significantly affect the rate of the COVID-19 transmission. Through this study, we have shown that connections between networks can be analyzed in terms of the centrality and eigenvectors of a matrix. This study recommends that containment effort to prevent exponential rise the COVID-19 cases must be focused on Cebu City for the following reasons; (1) Cebu City belongs to the cluster with the highest modularity class; (2) Cebu city has the highest degree centrality; and Cebu City belongs to the cluster with the highest eigenvector centrality.

Published

2020-11-01

Issue

Section

Articles