Enormous amount of available data in relational database (RDB) format creates a demand for automatic transforming them into OWL ontology to reuse in the Semantic Web. Many approaches have been proposed, however, most of them simply generate output ontology as the same flat structure with the original database and result in redundancy of ontology data. As an attempt to resolve the redundant problem, we propose a novel approach to generate OWL ontology from relational database while focusing on the similarity measure of duplicate attributes in relational tables. Experimental results show that the proposed method reliably predicts semantic similarity of duplicate columns and produces a better-quality OWL ontology.
Research on Learning OWL Ontology from Relational Database