Modeling the Spread and Control of a Contagion on Multiplex Social Networks

As social networks are composed of populations and contacts among their members, epidemiologists have used networks to model contagions spreading through populations. Such networks, however, do not reflect the dynamic nature of real social networks. To solve this problem, I modeled a contagion on a multiplex network comprised of individual social networks. This implementation allowed a contagion to spread via the edges of two individual networks in alternation. I then considered the effects of vaccination on the transmission of a contagion. To reflect limited vaccine supply in the real world, I used three different selection criteria to determine immunized nodes. The results showed that vaccination based on only one of the two graphs in the multiplex was largely ineffective, indicating that an alternative method for vaccination is necessary. This work demonstrates that multiplex networks can be used to represent a greater variety of social dynamics and simulate disease outbreaks with them. This research has applications in the public health sector, where epidemiologists and public health researchers, using the knowledge of their population's structure, can formulate strategies on how to best protect their constituents. Finally, since the model applies to the propagation of a contagion on networks, it could also find use in simulating the spread of social ideas, rumors, and unrest. This research was supported by the Center for Computational Epidemiology and Response Analysis, directed by Dr. Armin R. Mikler.