Research data related to Document-based Ontology (DbO) sets for the second experiment - 1000 in each corpus