Semantic Web Dog Food

Presented at: 7th International Semantic Web Conference (ISWC2008)

by Juan F. Sequeda, Syed Tirmizi, Daniel Miranker

The vision of the Semantic Web is to create a web of data with well-defined meaning. Most data in the current web is managed by relational databases. Thus, it is imperative for the Semantic Web community to offer easily implemented solutions to bridging relational database content and RDF. Direct mappings means to use the SQL schema to create an OWL ontology and use it to represent the data in RDF. Direct mapping methods have an advantage that they are, intrinsically, automated. If a SQL-schema was created using contemporary model-driven software engineering tools, the resulting OWL ontology can be semantically rich. However, few SQL databases are so developed and therefore semantically weak. We suggest that direct-mapping methods can be integrated by a refinement process. We propose a two step bootstrapping architecture of integrating relational databases with the Semantic Web by first generating a “database-derived putative ontology” and second, refining the putative ontology with a domain ontology and database individuals.

Keywords: Direct Mapping, Ontology, Relational Databases, Semantic Web

Resource URI on the dog food server: http://data.semanticweb.org/conference/iswc/2008/paper/poster_demo/74

Explore this resource elsewhere:

• on Snorql
• on Sindice
• on dotAC Explorer