Dynamics of social networks

TL;DR

This paper explores how simple local rules in social network models can lead to complex global structures like small-world and scale-free properties, matching real-world social network data.

Contribution

It introduces a basic dynamic model that reproduces key features of social networks, including clustering and degree distributions, advancing understanding of their formation.

Findings

Model produces highly clustered networks with small-world properties.

Results align with empirical data from coauthorship networks.

Network degree distributions are both scale-free and exponential.

Abstract

Complex networks as the World Wide Web, the web of human sexual contacts or criminal networks often do not have an engineered architecture but instead are self-organized by the actions of a large number of individuals. From these local interactions non-trivial global phenomena can emerge as small-world properties or scale-free degree distributions. A simple model for the evolution of acquaintance networks highlights the essential dynamical ingredients necessary to obtain such complex network structures. The model generates highly clustered networks with small average path lengths and scale-free as well as exponential degree distributions. It compares well with experimental data of social networks, as for example coauthorship networks in high energy physics.