The impact of interaction and algorithm choice on identified communities

In social networks, nodes are organized into densely linked communities where edges appear among the nodes with high concentration. Identifying communities has proven to be a challenging task due to various community definitions/algorithms and also due to the lack of “ground truth” for reference and evaluation. These communities not only differ due to various definitions but also can be affected by the type of interactions modeled in the network, which lead to different social groups. We are interested in exploring and studying the concept of partial network views, which is based on multiple types of interactions. An Enron email network is used to conduct our experiments. In this paper, we explore the mutual impact of selecting different views extracted from the same network and their interplay with various community detection algorithms to measure the change and the level of realism of the structure for non-overlapping communities. To better understand this, we assess the agreement of partitions by evaluating the partitioning quality (performance) and finding the similarity between algorithms. The results demonstrate that the topological properties of communities and the performance of algorithms are equivalent to each other. Both of them are affected by the type of interaction specified in each view. Some network views appeared to have more interesting communities than other views, thus, might help to approach a relatively informative and logic “ground truth” for communities.

[1]  E A Leicht,et al.  Community structure in directed networks. , 2007, Physical review letters.

[2]  A Díaz-Guilera,et al.  Self-similar community structure in a network of human interactions. , 2003, Physical review. E, Statistical, nonlinear, and soft matter physics.

[3]  Jari Saramäki,et al.  Characterizing the Community Structure of Complex Networks , 2010, PloS one.

[4]  Jafar Adibi,et al.  The Enron Email Dataset Database Schema and Brief Statistical Report , 2004 .

[5]  William M. Rand,et al.  Objective Criteria for the Evaluation of Clustering Methods , 1971 .

[6]  Stijn van Dongen,et al.  Graph Clustering Via a Discrete Uncoupling Process , 2008, SIAM J. Matrix Anal. Appl..

[7]  Martin Rosvall,et al.  Maps of random walks on complex networks reveal community structure , 2007, Proceedings of the National Academy of Sciences.

[8]  Martin Rosvall,et al.  An information-theoretic framework for resolving community structure in complex networks , 2007, Proceedings of the National Academy of Sciences.

[9]  L. Hubert,et al.  Comparing partitions , 1985 .

[10]  M E J Newman,et al.  Modularity and community structure in networks. , 2006, Proceedings of the National Academy of Sciences of the United States of America.

[11]  M E J Newman,et al.  Fast algorithm for detecting community structure in networks. , 2003, Physical review. E, Statistical, nonlinear, and soft matter physics.

[12]  Matthieu Latapy,et al.  Computing Communities in Large Networks Using Random Walks , 2004, J. Graph Algorithms Appl..

[13]  M. A. Munoz,et al.  Improved spectral algorithm for the detection of network communities , 2005 .

[14]  Leto Peel,et al.  Estimating network parameters for selecting community detection algorithms , 2010, 2010 13th International Conference on Information Fusion.

[15]  Jean-Loup Guillaume,et al.  Fast unfolding of communities in large networks , 2008, 0803.0476.

[16]  Alan Kai-Hau Yeung,et al.  Similarity between community structures of different online social networks and its impact on underlying community detection , 2015, Commun. Nonlinear Sci. Numer. Simul..

[17]  Jure Leskovec,et al.  Empirical comparison of algorithms for network community detection , 2010, WWW '10.

[18]  Mohammed J. Zaki,et al.  Is There a Best Quality Metric for Graph Clusters? , 2011, ECML/PKDD.

[19]  J. Reichardt,et al.  Statistical mechanics of community detection. , 2006, Physical review. E, Statistical, nonlinear, and soft matter physics.

[20]  M E J Newman,et al.  Community structure in social and biological networks , 2001, Proceedings of the National Academy of Sciences of the United States of America.

[21]  Réka Albert,et al.  Near linear time algorithm to detect community structures in large-scale networks. , 2007, Physical review. E, Statistical, nonlinear, and soft matter physics.

[22]  T. Vicsek,et al.  Uncovering the overlapping community structure of complex networks in nature and society , 2005, Nature.

[23]  F. Radicchi,et al.  Benchmark graphs for testing community detection algorithms. , 2008, Physical review. E, Statistical, nonlinear, and soft matter physics.

[24]  Hocine Cherifi,et al.  Comparative evaluation of community detection algorithms: a topological approach , 2012, ArXiv.

[25]  M. Newman,et al.  Finding community structure in networks using the eigenvectors of matrices. , 2006, Physical review. E, Statistical, nonlinear, and soft matter physics.