Unifying the object-oriented paradigm with semantic data models.

摘要

The object-oriented paradigm can be used to model behavior, and to a lesser extent, the structure of a problem domain. Semantic data models describe structure and semantics. This paper unifies the behavioral focus of the object-oriented paradigm with the structural and semantic focus of semantic data models. It presents abstractions to model static and derived data, composite objects, part hierarchies, semantic constraints, and abstractions for identifying behavior. The abstractions keep the model close to the problem domain, are independent of language features, and can be translated into object-oriented, relational or network implementations. This paper makes three principal contributions. First, a comprehensive set of data structuring abstractions is described. Second, semantic constraints inherent in the graphical representation of the abstractions are identified. Third, abstractions for identifying behavior are described. 14 refs., 17 figs.