Inclusion Dependencies

Abstract

The evaluation of conceptual schemes of actual databases may result in the discovery of inclusion dependencies. An inclusion dependency is defined as the existence of attributes in a table whose values must be a subset of the values of attributes in another table. When the latter set conforms a key for its table, the inclusion dependency is key-based. Key-based inclusion dependencies are fully enforced by most current database systems. On the contrary, if the second set is not the key of the relation, the inclusion dependency is non-key-based. This kind of inclusion dependency is completely disregarded by actual systems, obliging the users to manage them via special-case code or triggers. This implies an excessive effort to maintain integrity and develop applications, among other inconveniences. The chapter goal is to give a heuristics to redesign the conceptual schema. This is based on the identification of hidden business rules and the conversion of non-key inclusion dependencies into key-based ones.