METHOD AND STRUCTURE FOR CLUSTERING DATABASE TABLES INTO CLASSES AND PRESENTING EACH CLASS AS AN E-R MODEL

A reverse engineering technology is capable of automatically reversing from a physical data model comprising a relational or 4GL data model of an existing application database into a logical data model comprising classes and an entity-relationship (E-R) model of each class, the classes representing the different types of high-level objects that make up the application. This logical data model allows users to have a much clearer understanding of their application database thereby increasing their productivity and ease of maintenance of their existing applications. This is made possible by the method and structure of this invention first capturing the application semantics from the physical data model to provide a set of derived semantics. The derived semantics are then used to identify classes and derive the class definitions which are then used to derive the E-R model of each class. In addition, the derived semantics can be easily updated by the method and structure of this invention when the physical data model is changed.