© Springer-Verlag Berlin Heidelberg 2008 Translating SQL Applications to the Semantic Web

Abstract. The content of most Web pages is dynamically derived from an un-derlying relational database. Thus, the success of the Semantic Web hinges on enabling access to relational databases and their content by semantic methods. We define a system for automatic transformation of SQL DDL schemas into OWL DL ontologies. This system goes further than earlier efforts in that the en-tire system is expressed in first-order logic. We leverage the formal approach to show the system is complete with respect to a space of the possible relations that can be formed among relational tables as a consequence of primary and foreign key combinations. The full set of transformation rules is stratified, thus the system can be executed directly by a Datalog interpreter.