EPIDEMIC SPREAD ANALYSIS IN SOCIAL COMMUNICATION NETWORKS WITH SIR MODEL

Abstract

Compartmental mathematical models are frequently used in epidemiology. These types of models rely on some assumptions, such as the homogeneity of the society and the equal contact ratio of everyone, to model real-life events mathematically. In real life, due to the heterogeneous nature of the social network that constitutes society, the contact rates and contact times of individuals vary. In sudden and new types of epidemics, solutions such as vaccines to slow down or end epidemics may be limited. In such cases, it becomes more important to use limited resources with maximum efficiency. In this study, the estimation results of disease spread in homogeneous and heterogeneous population structures were compared using the SIR compartment model. The dataset obtained from the science gallery in Dublin in 2009 was used to illustrate the heterogeneous community structure in real life. In the exhibition, the spread of the disease was simulated when individuals with different degrees of centrality in the network formed by the visitors who made face-to-face contacts were immunized. When the results obtained are compared, in the case of vaccination of individuals with high betweenness centrality, the spread of infection occurs 14,39% less than the homogeneous network structure accepted in SIR models.