The paper describes a general framework allowing the specification, comparison and conversion of most information and data models proposed in Information System design, be they conceptual or at the DBMS level. That framework is based on an extended N1NF relational model that includes the concept of entity domain. The analysis of the constructs available in current information and data models, and particularly ER models, demonstrates the ability of the proposed framework to express these constructs without loss of semantics. The GER is a large subset of this extended relational model that can uniquely represent the concept of most current entity-based and object-based models, while being more regular, less complex and more powerful than their union. By further augmenting the GER with two simple logical implementation constructs, we extend it to the expression of DBMS-dependent models. The GER model has been used as the basis of database design methodologies and tools.
|titre||Information System Concepts: an in-depth analysis|
|Sous-titre||Proc. of the IFIP WG 8.1 Conf. on Information System Concepts|
|Nombre de pages||30|
|Etat de la publication||Publié - 1989|