Time-varying networks approach to social dynamics: From individual to collective behavior

TL;DR

This thesis investigates the importance of the temporal dimension in social networks by introducing time-varying network formalism, analyzing empirical data, and exploring dynamical processes like spreading and random walks.

Contribution

It introduces a formalism for time-varying networks, analyzes empirical social data, and develops models and analytical tools for understanding dynamics on temporal networks.

Findings

Human contact networks exhibit heterogeneity and burstiness.

The activity-driven model captures key properties of social interactions.

Analytic methods can predict percolation and structural properties of temporal networks.

Abstract

In this thesis we contribute to the understanding of the pivotal role of the temporal dimension in networked social systems, previously neglected and now uncovered by the data revolution recently blossomed in this field. To this aim, we first introduce the time-varying networks formalism and analyze some empirical data of social dynamics, extensively used in the rest of the thesis. We discuss the structural and temporal properties of human contact networks, such as heterogeneity and burstiness of social interactions, and we present a simple model, rooted on social attractiveness, able to reproduce them. We then explore the behavior of dynamical processes running on top of temporal networks, constituted by empirical face-to-face interactions, addressing in detail the fundamental cases of random walks and epidemic spreading. We also develop an analytic approach able to compute the structural and percolation properties of the activity driven model, aimed to describe a wide class of social interactions, driven by the activity of the individuals involved. Our contribution in the rapidly evolving framework of social time-varying networks opens interesting perspectives for future work, such as the study of the impact of the temporal dimension on multi-layered systems.