Concept-Oriented Model: Extending Objects with Identity, Hierarchies and Semantics

The concept-oriented data model (COM) is an emerging approach to data modeling which is based on three novel principles: duality, inclusion and order. These three structural principles provide a basis for modeling domain-specific identities, object hierarchies and data semantics. In this paper these core principles of COM are presented from the point of view of object data models (ODM). We describe the main data modeling construct, called concept, as well as two relations in which it participates: inclusion and partial order. Concepts generalize conventional classes by extending them with identity class. Inclusion relation generalizes inheritance by making objects elements of a hierarchy. We discuss what partial order is needed for and how it is used to solve typical data analysis tasks like logical navigation, multidimensional analysis and reasoning about data.