DIRECT MAPPING SQL DATABASES TO THE SEMANTIC WEB: A SURVEY

The Semantic Web anticipates rich, integrated access to a large number of information sources on the Internet. Since a large number of websites are backed by SQL databases, methods that automate the integration of these databases with the Semantic Web are crucial. This paper surveys methods used to create ontological descriptions of databases by examining the database schema, normally known as direct mapping methods. Compared to wrapping methods, which map databases to existing legacy ontologies, direct mapping methods have the advantage that they are intrinsically automated. In this review we first identify a correspondence between individual syntactic structures in SQL data description language (DDL) and layers of the Semantic Web stack, and we then characterize each research effort's transformation methods with respect to the expressive hierarchy of the Semantic Web languages themselves. Much of the prior work on direct mapping is expository in nature, relying heavily on examples. From this foundation, we define a unifying transformation system using transformation rules defined using first-order-logic. We show that the unified system is complete with respect to all possible associations formed by combinations of primary-key and foreign-key constructs. The unified transformation system has been implemented using the JESS rule engine, and has been made available to the community as a web-service.