Interpretable and reconfigurable clustering of document datasets by deriving word-based rules.