One of the most difficult issues in cartography is generalization. It is sometimes required to simplify the data in order to produce a map which visually adapts the purpose of the study also the statistics on this map should be developed in a way representing the main features. One of the most important steps in generalization is to aggregate similar phenomena to be represented as a whole phenomenon. In recent years, rapid development of multimedia technology has provided the opportunity to create and visualize the accurate 3D models of cities. This paper focuses on aggregation operator and tries to present and implement a method to aggregate 3D buildings in the process of generalization.