Clustering categorical data using coverage density

Abstract

In this paper, a new algorithm based on the idea of coverage density is proposed for clustering categorical data. It uses average coverage density as the global criterion function. Large sparse categorical databases can be clustered effectively by using this algorithm. It shows that the algorithm uses less memory and time by analyzing its time and space complexity. Experiments on two real datasets are carried out to illustrate the performance of the proposed algorithm. © Springer-Verlag Berlin Heidelberg 2005.