The advance of the Web has significantly and rapidly changed the way of information organization, sharing and distribution. The next generation of the web, the semantic web, seeks to make information more usable by machines by introducing a more rigorous structure based on ontologies. In this context we try to propose a novel and integrated approach for a semi-automated extraction of ontology-based semantic web from data-intensive web application and thus, make the web content machine-understandable. Our approach is based on the idea that semantics can be extracted by applying a reverse engineering technique on the structures and the instances of HTML-forms which are the most convenient interface to communicate with relational databases on the current dataintensive web application. This semantics is exploited to produce over several steps, a personalised ontology.