-->

Resources

Edge clustering community detection algorithm
In the paper Defining and identifying communities in networks we proposed a community detection algorithm based on the edge clustering coefficient.The code of the algorithm can be downloaded here. Instructions on how to use the code are included in the package.
This software can be cited as Defining and identifying communities in networks
F. Radicchi, C. Castellano, F. Cecconi, V. Loreto and D. Parisi
Proc. Natl. Acad. Sci. USA 101, 2658-2663 (2004) [pdf].
Benchmark for community detection algorithms
In the paper Benchmark graphs for testing community detection algorithms we introduced a new class of artificial networks that pose a far harder test to community detection algorithms. The new benchmark is an extension of the benchmark by Girvan and Newman. In the latter, the nodes have the same degree and the communities have equal size. Here, the distributions of nodes’ degree and community size are power laws, with tunable exponents. The code to build the new benchmark graphs can be downloaded here. Instructions on how to use the code are included in the package.
This software can be cited as Benchmark graphs for testing community detection algorithms, A. Lancichinetti, S. Fortunato and F. Radicchi, Phys. Rev. E 78, 046110 (2008) [pdf].
Ebay dataset
The dataset consists of two files
1 Activity This file has 733335 lines. Each of them corresponds to a user and contains a variable number of columns. The first column reports the number of pairs between consecutive messages sent by the user. Then for each of these pairs is reported the time difference (resolution in hours). The last column contains the sum of all inter-event periods of time. The size of the file is 121Mb (384Mb unzipped).
2 Replies This file contains 6511710 pairs message/reply. Each line contains three columns: the first two columns report the ids of the users, while the third contains the time difference (resolution in hours) between message and reply. The size of the file is 32Mb (105Mb unzipped).
These data can be cited as Human Activity in the Web, F. Radicchi, Phys. Rev. E 80, 026118 (2009) [pdf].
Phys Author Rank Algorithm is a website where physicists can check the evolution of their own scientific rank. Scientific rank is calculated using the Science Author Rank Algorithm on a weighted author citation network.
This ranking procedure can be cited as Diffusion of scientific credits and the ranking of scientists, F. Radicchi, S. Fortunato, B. Markines and A. Vespignani, Phys. Rev. E 80, 056103 (2009) [pdf]
Statistical significance of communities in networks
This software allows to calculate the statistical significance of communities in networks. You may download the source code (compilation requires GSL) or test the significance of your communities by simply using our on-line tool.
This software can be cited as Statistical significance of communities in networks, A. Lancichinetti, F. Radicchi and J.J. Ramasco, arXiv: 0907.3708 (2009) [pdf]