This document summarizes incremental conceptual clustering, specifically Fisher's Cobweb algorithm. It begins by discussing hierarchical clustering approaches like single and complete link clustering. It then focuses on Cobweb, which uses a hill climbing search strategy to incrementally construct concept hierarchies from observations. Cobweb's category utility function measures intra-class similarity and inter-class predictiveness to evaluate candidate clusterings. The algorithm can classify observations into existing clusters, create new clusters, or merge and split clusters to construct the hierarchy.