Critical Transitions in Social Network Activity

TL;DR

This paper investigates whether early warning signs like variance and autocorrelation, known from other complex systems, can be detected in social network data before known critical events occur.

Contribution

It provides the first empirical evidence that stochastic warning signs are observable in social networks prior to known events, bridging theory and social data analysis.

Findings

Variance and autocorrelation increase before known social events

Empirical support for warning signs in social network data

Establishes a basis for further theoretical-social system studies

Abstract

A large variety of complex systems in ecology, climate science, biomedicine and engineering have been observed to exhibit tipping points, where the internal dynamical state of the system abruptly changes. For example, such critical transitions may result in the sudden change of ecological environments and climate conditions. Data and models suggest that detectable warning signs may precede some of these drastic events. This view is also corroborated by abstract mathematical theory for generic bifurcations in stochastic multi-scale systems. Whether the stochastic scaling laws used as warning signs are also present in social networks that anticipate a-priori {\it unknown} events in society is an exciting open problem, to which at present only highly speculative answers can be given. Here, we instead provide a first step towards tackling this formidable question by focusing on a-priori {\it known} events and analyzing a social network data set with a focus on classical variance and autocorrelation warning signs. Our results thus pertain to one absolutely fundamental question: Can the stochastic warning signs known from other areas also be detected in large-scale social network data? We answer this question affirmatively as we find that several a-priori known events are preceded by variance and autocorrelation growth. Our findings thus clearly establish the necessary starting point to further investigate the relation between abstract mathematical theory and various classes of critical transitions in social networks.