Aaron McDaid

The use of modularity for community finding is based on what I would call a minimum likelihood approach. The idea behind modularity involves creating a model that is as wrong as possible, called a null model, then find the parameters for which the fit between the data and the model is a bad as possible.