Quantifying dynamical spillover in co-evolving multiplex networks

TL;DR

This paper introduces a Markov chain-based null model to quantify and analyze dynamical spillover in co-evolving multiplex networks, revealing correlated edge dynamics in real-world systems like international relations and open source software development.

Contribution

The study presents a novel method to measure edge correlation in multiplex networks using longitudinal data, enabling detection of dynamical spillover effects.

Findings

Existence of dynamical spillover in real-world networks

Correlated edge formation and deletion across layers

Insights into causal pathways of network evolution

Abstract

Multiplex networks (a system of multiple networks that have different types of links but share a common set of nodes) arise naturally in a wide spectrum of fields. Theoretical studies show that in such multiplex networks, correlated edge dynamics between the layers can have a profound effect on dynamical processes. However, how to extract the correlations from real-world systems is an outstanding challenge. Here we provide a null model based on Markov chains to quantify correlations in edge dynamics found in longitudinal data of multiplex networks. We use this approach on two different data sets: the network of trade and alliances between nation states, and the email and co-commit networks between developers of open source software. We establish the existence of "dynamical spillover" showing the correlated formation (or deletion) of edges of different types as the system evolves. The details of the dynamics over time provide insight into potential causal pathways.