Nowadays data management systems and knowledge management systems tend to converge. Our work is part of the OSIRIS data and knowledge management system. This system is centered on views which are said object-preserving (i.e. they do not create new objects). Our work led to the construction of a model for the OSIRIS System inspired from those of Description Logics. In this model we have defined the main functionalities of databases (insert, delete, modify, select) and description logics (subsumption). In this manner we show that there is a close relationship between subsumption calculus and query evaluation with the notion of interpretation. The data stored in a database can be seen as an interpretation (same meaning as in first order logic) of the set of axioms constituted by the databases’ scheme. Typically a database manipulates what is true in a single interpretation, a knowledge base what is true in all possible interpretations.