Ontology learning from data models: Advance and the requirement for a database preparation and enrichment process

Ontology learning field provides automatic and semi-automatic approaches for ontology generation. This process aims to overcome the manual knowledge acquisition process that is difficult, expensive and time-consuming. Several approaches have been emerged for this purpose. However, they are not practical on real-world databases and they are far from capturing all application semantics. This paper gives, first, an overview of the ontology learning process and the peculiarities of relational models. Then, it discusses the advance in this filed according to a developed framework before presenting two main lacks: the requirement of a database preparation process and an enrichment process.