Topological Decomposition and Heuristics for High Speed Clustering of Complex Networks
Abstract
With the exponential growth in the size of data and networks, development
of new and fast techniques to analyze and explore these networks is
becoming a necessity. Moreover the emergence of scale free and small world
properties in real world networks has stimulated lots of activity in the field of
network analysis and data mining. Clustering remains a fundamental technique
to explore and organize these networks. A challenging problem is to find a clustering
algorithm that works well in terms of clustering quality and is efficient in
terms of time complexity.
In this paper, we propose a fast clustering algorithm which combines some
heuristics with a Topological Decomposition to obtain a clustering. The algorithm
which we call Topological Decomposition and Heuristics for Clustering
(TDHC) is highly efficient in terms of asymptotic time complexity as compared
to other existing algorithms in the literature. We also introduce a number of
Heuristics to complement the clustering algorithm which increases the speed of
the clustering process maintaining the high quality of clustering. We show the
effectiveness of the proposed clustering method on different real world data sets
and compare its results with well known clustering algorithms.