An algorithm is described which constructs a hierarchical taxonomy over object sets. The algorithm falls within the purview of conceptual clustering, and is computationally more efficient than conceptual clustering algorithms currently reported in the literature.
