This talk will present a meta-modeling approach to formalize a set of metrics suited for object-relational databases. Those metrics are expected to help database designers in their activity, offering a mechanism for comparing and selecting among possible schema designs the one with more quality. Formalization of metrics is important for avoiding misinterpretation of definitions, while producing metrics collection tools. The metamodel used for formalization purposes is represented as a UML class diagram and is based on the newest version of the ISO SQL:2003 standard. The formalizing technique uses OCL, the constraint language of UML, for expressing the metrics, which in turn allows automating metrics collection. Some extracts from an object-relational database schema will be used to illustrate the metrics collection process using the proposed approach.
