Automatic Generation of Ontology from Data Models: A Practical Evaluation of Existing Approaches

Nowadays, ontology as a knowledge sharing approach plays an important role in semantic interoperability of enterprise applications (EAs). However, the manual process of ontology construction requires deep understanding of the domain. This approach is difficult, expensive and time-consuming. To overcome the knowledge acquisition bottleneck, ontology learning field aims to provide automatic and semi-automatic approaches for ontology generation. Several approaches have been emerged for this purpose. In this paper, we present a practical study of methods that take data models as input to the learning process. The main contributions of this work are: (i) the evaluation of the availability of existing approaches for (semi-)automatic generation of ontology from data models; (ii) the evaluation of tools according to their operability; and (iii) the evaluation of the resulting ontologies to assess their quality in supporting semantic interoperability. Our goal through this study is to find a response to the question: Is there a tool that extracts (semi-) automatically an application ontology from data models, intended for use in semantic interoperability?.