Dynamics of social networks

Abstract Complex networks such 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 nontrivial 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. © 2003 Wiley Periodicals, Inc.