Abstract: In this research work, we deliberate how ontologies can be constructed by mining information from databases and Web documents in web service description language. This is a stimulating task, because information mining is typically a noisy effort whereas ontologies generally involve hygienic and crispy data. This means that the mined information has to be cleaned, disambiguated and prepared logically stable to certain step. We discuss approaches that extract ontology in this spirit from the relational databases and also present methodologies that objective to extract ontology from documents or by extension from the entire Web then implemented with the improved enhanced traversal algorithm on the databases such as bank and hospital then obtain better accuracy performance. We will show that information mining and ontology creation can arrive into a successful support loop, where additional mined information pointers to a bigger ontology which helps mining more information.

Keywords: Ontology, Semantic Web Services, Bootstrapping, Improved ET Algorithm, WSDL.