The Scaling laws of Spatial Structure in Social Networks

TL;DR

This paper investigates the universal spatial structure properties of social networks, revealing a scale-invariant geographic distance distribution that enhances information searchability and offers deep insights into human interaction patterns.

Contribution

It uncovers a scale-invariant property of geographic distances in social networks, contributing to understanding their universal structural features.

Findings

Geographic distance distribution follows Pr(d) ∝ d^{-1}

Spatial structure is scale-invariant across social networks

This property benefits information searching in social contexts

Abstract

Social network structure is very important for understanding human information diffusing, cooperating and competing patterns. It can bring us with some deep insights about how people affect each other. As a part of complex networks, social networks have been studied extensively. Many important universal properties with which we are quite familiar have been recovered, such as scale free degree distribution, small world, community structure, self-similarity and navigability. According to some empirical investigations, we conclude that our social network also possesses another important universal property. The spatial structure of social network is scale invariable. The distribution of geographic distance between friendship is about $Pr(d)\propto d^{-1}$ which is harmonious with navigability. More importantly, from the perspective of searching information, this kind of property can benefit individuals most.