Vivaldi: a decentralized network coordinate system
Proceedings of the 2004 conference on Applications, technologies, architectures, and protocols for computer communications
Network Traffic Classification Using K-means Clustering
IMSCCS '07 Proceedings of the Second International Multi-Symposiums on Computer and Computational Sciences
Graph Clustering Via a Discrete Uncoupling Process
SIAM Journal on Matrix Analysis and Applications
CDNs Content Outsourcing via Generalized Communities
IEEE Transactions on Knowledge and Data Engineering
Approximate Distributed K-Means Clustering over a Peer-to-Peer Network
IEEE Transactions on Knowledge and Data Engineering
ISCC '11 Proceedings of the 2011 IEEE Symposium on Computers and Communications
Computer Science Review
| Hi-index | 0.00 |
A fundamental problem in networking and computing is community detection. Various applications like finding web communities, uncovering the structure of social networks, or even analyzing a graph's structure to uncover Internet attacks are just some of the applications for which community detection is important. In this paper, we propose an algorithm that finds the entire community structure of a network, represented by an undirected, unweighted graph, based on local interactions between neighboring nodes and on an unsupervised centralized clustering algorithm. The novelty of the proposed approach is the fact that the algorithm is based on the use of network coordinates computed by a distributed algorithm. Experimental results and comparisons with the Lancichinetti et al. method (Phys. Rev. E 80(5 Pt 2), 056117, 2009; New J. Phys. 11(3), 033015, 2009) are presented for a variety of benchmark graphs with known community structure, derived by varying a number of graph parameters. Emphasis is given on benchmark graphs with significant variations in the size of their communities. Further experimental results are presented for two real dataset graphs, namely the Enron, and the Epinions graphs, from SNAP, the Stanford Large Network Dataset Collection. The experimental results demonstrate the high performance of our algorithm in terms of accuracy to detect communities, and its computational efficiency.