Social network structures that achieve both infection control and knowledge discovery efficiency

Abstract

With the global outbreak of COVID-19, it is necessary to acquire social network structures that can achieve both infectious disease control and socio-economic activities. The purpose of this article is to exploratively reveal such network structures and to realize a method to obtain them. Our method consists of two steps. First, in order to survey which networks can achieve both two objectives, we conduct two simulations that model these phenomena on various networks, and investigate which network features are strongly related to these efficiencies. Next, we search for the optimal network structure using the genetic algorithm, whose evaluation function is set to the obtained features. From the results of the survey of network features, it was found that large clustering coefficients and small maximum eigenvalues of the adjacency matrix are important for the two objectives. For the problem set up to simulate actual changes in social network structure for infection control of COVID-19, optimization of these network features using the genetic algorithm was performed, and it was confirmed that our method can obtain network structures that achieve the two objectives in some cases.