An Algorithm for Mining Weighted Dense Maximal 1-Complete Regions

We propose a new search algorithm for a special type of subspace clusters, called maximal 1-complete regions, from high dimensional binary valued datasets. Our algorithm is suitable for dense datasets, where the number of maximal 1-complete regions is much larger than the number of objects in the datasets. Unlike other algorithms that find clusters only in relatively dense subspaces, our algorithm finds clusters in all subspaces. We introduce the concept of weighted density in order to find interesting clusters in relatively sparse subspaces. Experimental results show that our algorithm is very efficient, and uses much less memory than other algorithms.