Classifying schematic and data heterogeneity in multidatabase systems

A complete framework for enumerating and classifying the types of multidatabase system (MDBS) structural and representational discrepancies is developed. The framework is structured according to a relational database schema and is both practical and complete. It was used to build the UniSQL/M commercial multidatabase system. This MDBS was built over Structured-Query-Language-based relational database systems and a unified relational and object-oriented database system named UniSQL/X. However, the results are substantially applicable to heterogeneous database systems that use a nonrelational data model (for example, an object-oriented data model) as the common data model and allow the formulation of queries directly against the component database schemas.>