30.08.2016 10:29 Age: 1 year
Category: Publications
By: Jean-Gabriel Young

New paper in Phys. Rev. E

The density of communities scales inversely with the their size, regardless of the algorithm used.

Social networks encode the interactions of the elements of a social system; one thinks, for instance, of Facebook, where individuals interact through friendship relations. A striking feature of most social networks is their modularity, a feature implying that individuals are not connected at random but rather that connections are the reflection of social communities. Structurally, small communities---such as group of friends or nuclear families---are usually tightly connected groups, whereas larger communities---such as workplaces or schools---are loosely connected group of individuals.

In our latest paper, titled "Growing networks of overlapping communities with internal structure", we argue that this difference of connection density is the result of the continuous growth of social networks. We support this proposition with empirical evidence, and present a general recipe to realistically model real social networks. Our approach offers a natural and dynamical explanation of the Dunbar number, i.e., an upper bound on the number of connection that individuals can sustain in a social network.

The full text is available on the Publications page.