An algorithm for network community structure determination by surprise

Graphs representing real world systems may be studied from their underlying community structure. A community in a network is an intuitive idea for which there is no consensus on its objective mathematical definition. The most used metric in order to detect communities is the modularity, though many disadvantages of this parameter have already been noticed in the literature. In this work, we present a new approach based on a different metric: the surprise. Moreover, the biases of different community detection algorithms and benchmark networks are thoroughly studied, identified and commented about.