Clustering massive categorical data with class association rules

Clustering algorithms partition data sets into of objects such that the pairwise similarity objects within the same cluster is higher than assigned to different clusters. Defining a measure becomes challenging in the of categorical data and affects the quality meaningfulness of the clusters formed. , the curse of dimensionality diminishes robustness of such measures. This paper introduces (Supervised Clustering with Association Rules) nontraditional algorithm for clustering massive high categorical data. SCAR is robust to the of dimensionality, it relies on association rules an intuitive way to evaluate the similarity between objects and group them.