Font Size: 

DESPOTA is a clustering method that cuts the tree branches at various heights to find the best division among those that can be achieved from the hierarchical clustering tree through the use of a permutation test at each node. In order to reduce the computational cost and increase the applicability of DESPOTA to huge data sets, the present study suggests two improvements to the DESPOTA original implementation that combine aggregation with either splitting or partitioning approaches. A dataset of the Italian universities' five-year periodical accreditation by the Italian national agency (ANVUR) is used to test the suggested approach.