Reasoning about ER models in a deductive environment

摘要

In this paper we present an approach to represent schema information, application data and integrity constraints as a logic program in form of Datalog. The schema information is supplied as an enhanced entity relationship (EER) model which is transformed by a one-to-one mapping into a set of ground facts. The application data corresponding to the schema is also represented by ground facts in a single table. In order to check whether the application data conforms to the given schema, generic functional and inclusion dependencies are introduced, which can be mapped into concrete dependencies using the schema information. The concrete dependencies are used for checking the consistency between application data and the schema. The formulation of the constraints based on functional and inclusion dependencies led to a small extension of the EER model by allowing identifying attributes in relationship types. This extension leads to both simpler constraints and simpler EER application models.Furthermore, we provide a meta EER model which can be used to check whether the application EER model is a structural valid EER model. Any application EER diagram is an instance of the meta EER diagram which can be specified using the proposed application data representation. The same integrity rules can be used to check the conformance between the application data and the application EER diagram, the meta EER diagram and the application EER diagram, and finally it can be used to check the meta EER model itself.