Elements of information theory
Elements of information theory
Inferring Web communities from link topology
Proceedings of the ninth ACM conference on Hypertext and hypermedia : links, objects, time and space---structure in hypermedia systems: links, objects, time and space---structure in hypermedia systems
Diffusion Kernels on Graphs and Other Discrete Input Spaces
ICML '02 Proceedings of the Nineteenth International Conference on Machine Learning
| Hi-index | 0.00 |
In the real world, many systems can be represented as a network, in which the nodes denote the objects of interest and the edges describe the relations between them, such as telecommunication networks, power grid networks, and email communication networks These complex networks have been revealed to possess many common statistical properties such as scale-free nature and small-world property In addition, modularity or community structure is another important characteristic of complex networks Identifying modular structure can help us understand the function of networks In this paper, we introduce a method based on information-theoretic clustering for finding communities/modules in complex networks This method is robust to the feature representation of networks Moreover, unlike most existing algorithms, this method does not need to search the number of communities in a network and can determine it automatically We apply this method to several well-studied networks including a large-scale email communication network and the computational results demonstrate its effectiveness.