Dynamical processes on networks occur all around us: neural computations in the brain, friendship interactions on social networks, genetic regulation in metabolic pathways, and data transmission over the internet, to name a few examples studied under the umbrella of complex systems. The ubiquity of measurements from these systems ('big data') and the advent of tools from modern non-parametric statistics allow us to ask and answer new questions about how these systems behave and evolve over time.
My current work focuses on the interaction of and between users on Twitter. In collaboration with the Social Media Micro-Modeling (SM3) research group at the University of Maryland, I am undertaking an empirical study of the behavior of fifteen thousand (15K) users on Twitter over a three month period. By inferring models for the dynamics of these users, we can assign 'barcodes' to the users as to their apparent randomness and complexity. In this way, we compress second-resolution behavior over the course of three months into a small collection of numbers that allow us to classify users based on various typical barcodes.