Ontology-driven relational query formulation using the semantic and assertional capabilities of OWL-DL

摘要

This work investigates the extent to which domain knowledge, expressed in a domain ontology, can assist end-users in formulating relational queries that can be executed over a complex relational database. In this regard, an ontology-driven query formulation architectural framework has been devised, namely OntoQF, that implements a two-phased approach – the pre-processing and translation phases. In the pre-processing phase, a new database-to-ontology transformation approach has been synthesised where domain ontology is populated and enriched with problem domain concepts and semantic relationships specified using OWL-DL. Once domain ontology has been formulated, end-users can write sophisticated ontology-based queries that are then translated, in the translation phase, into the corresponding relational query statements. In order to validate the correctness of translating single or multiple OWL-DL constructs into their corresponding relational ones, a set of test cases have been derived from the medical domain. Our results demonstrated that the OntoQF framework enriches domain ontology and its associated algorithms drive the process of relational query formulation without the need to both replicate transactional data into the associated domain ontology and have knowledge of the underlying database schema.