%0 Generic %1 citeulike:554 %A Orponen, Pekka %A Schaeffer, Satu E. %D 2004 %K characterization, web-graph %T Efficient Algorithms for Sampling and Clustering of Large Nonuniform Networks %U http://arxiv.org/abs/cond-mat/0406048 %X We propose efficient algorithms for two key tasks in the analysis of large nonuniform networks: uniform node sampling and cluster detection. Our sampling technique is based on augmenting a simple, but slowly mixing uniform MCMC sampler with a regular random walk in order to speed up its convergence; however the combined MCMC chain is then only sampled when it is in its üniform sampling" mode.Our clustering algorithm determines the relevant neighbourhood of a given node u in the network by first estimating the Fiedler vector of a Dirichlet matrix with u fixed at zero potential, and then finding the neighbourhood of u that yields a minimal weighted Cheeger ratio, where the edge weights are determined by differences in the estimated node potentials. Both of our algorithms are based on local computations, i.e. operations on the full adjacency matrix of the network are not used. The algorithms are evaluated experimentally using three types of nonuniform networks: Dorogovtsev-Goltsev-Mendes "pseudofractal graphs", scientific collaboration networks, and randomised "caveman graphs".

@misc{citeulike:554, abstract = {We propose efficient algorithms for two key tasks in the analysis of large nonuniform networks: uniform node sampling and cluster detection. Our sampling technique is based on augmenting a simple, but slowly mixing uniform MCMC sampler with a regular random walk in order to speed up its convergence; however the combined MCMC chain is then only sampled when it is in its "uniform sampling" mode.Our clustering algorithm determines the relevant neighbourhood of a given node u in the network by first estimating the Fiedler vector of a Dirichlet matrix with u fixed at zero potential, and then finding the neighbourhood of u that yields a minimal weighted Cheeger ratio, where the edge weights are determined by differences in the estimated node potentials. Both of our algorithms are based on local computations, i.e. operations on the full adjacency matrix of the network are not used. The algorithms are evaluated experimentally using three types of nonuniform networks: Dorogovtsev-Goltsev-Mendes "pseudofractal graphs", scientific collaboration networks, and randomised "caveman graphs".}, added-at = {2009-08-06T15:16:38.000+0200}, archiveprefix = {arXiv}, author = {Orponen, Pekka and Schaeffer, Satu E.}, biburl = {https://www.bibsonomy.org/bibtex/2f41735a3404960815d8133a3cce45efa/chato}, citeulike-article-id = {554}, citeulike-linkout-0 = {http://arxiv.org/abs/cond-mat/0406048}, citeulike-linkout-1 = {http://arxiv.org/pdf/cond-mat/0406048}, eprint = {cond-mat/0406048}, interhash = {ac2d020cfd53131d31fcd6d272699106}, intrahash = {f41735a3404960815d8133a3cce45efa}, keywords = {characterization, web-graph}, month = {June}, posted-at = {2007-02-15 02:36:23}, priority = {0}, timestamp = {2009-08-06T15:16:49.000+0200}, title = {Efficient Algorithms for Sampling and Clustering of Large Nonuniform Networks}, url = {http://arxiv.org/abs/cond-mat/0406048}, year = 2004 }