An Automatic Approach for Transforming Relational Databases into OWL Ontologies

Transforming relational databases (RDBs) into ontologies provides applications with the ability to rely on semantic representations of data. Representing data using RDBs is a useful mechanism for organizing and managing data, nevertheless, it lacks semantic representation capabilities. Existing relational-into-ontology transformation methods often struggle with handling complex database schemas, replicating constraints, and automating the process. This paper proposes a rule-based approach that addresses these shortcomings through a comprehensive set of transformation rules tailored to disparate relational elements and constraints. The proposed method classifies RDB data semantics and utilizes schema mapping rules and data analysis techniques to recover implicit semantics that may be lost during data modelling. Furthermore, the proposed method has been implemented as a prototype and evaluated through a series of experiments on benchmark datasets, validating the effective preservation of meaning, integrity, and coverage of schemas. The obtained results demonstrate significant benefits of the proposed approach over existing methods in terms of expressiveness and automation of the transformation process.