The structure of growing social networks

The structure of growing social networks1,2 2,3 2Emily M. Jin , Michelle Girvan , and M. E. J. Newman1Engineering and Applied Sciences, Harvard University, Cambridge, MA 021382Santa Fe Institute, 1399 Hyde Park Road, Santa Fe NM 875013Department of Physics, Cornell University, Ithaca, NY 14853–2501We propose some simple models of the growth of social networks, based on three general principles:(1) meetings take place between pairs of individuals at a rate which is high if a pair has one ormore mutual friends and low otherwise; (2) acquaintances between pairs of individuals who rarelymeet decay over time; (3) there is an upper limit on the number of friendships an individual canmaintain. Using computer simulations, we find that models that incorporate all of these featuresreproduce many of the features of real social networks, including high levels of clustering or networktransitivity and strong community structure in which individuals have more links to others withintheir community than to individuals from other communities.I. INTRODUCTION of features of social networks—networks of acquaintancebetween people, for instance—and introduced a simplemodel of a (static) social network, which has since beenMany real-world systems take the form of networks—analyzed in depth in the physics literature [1,13–17]. So-nodes or “vertices” joined together by links or “edges.”cial networks also evolve, with new acquaintances form-Commonly cited examples include ...