In 2007, Falkowski et al. [1] presented DenGraph - a density based graph clustering algorithm for community detection. The algorithm overcomes the complexity problem, and therefore is able to handle also huge networks. Later on, the authors presented in [2] and [3] the even more efficient incremental version DenGraph-IO that adapts an existing clustering when the underlaying graph structure changes. In 2008, DenGraph was used for a temporal analysis of the Enron email data set. In 2009, we applied the proposed method on a music data set to analyse the music listen behaviour of users on the Last.fm platform. Based on the incremental algorithm we present the evolution of groups of users with similar music listening behavior over time in [4] and [5].
The experiments and the results of this dDM study are published as a part of Falkowski's PhD theses that is also available as book Falkowski[6].
