Majority of the search algorithms in microdata anonymization restrict themselves to a single privacy property and a single criteria to optimize. The solutions obtained are therefore of limited application since adherence to multiple privacy models is required to impede different forms of privacy attacks. Towards this end, we propose the concept of a property based generalization (PBG) to capture the non-dominance relationships that appear when multiple objectives are to be met in an anonymization process. We propose an evolutionary algorithm that can identify a representative subset of the set of PBGs for the purpose of decision making.