Purpose – The purpose of the research is to speed up the process of semantic web services by transformation of current Web services into semantic web services. This can be achieved by applying ontology learning techniques to automatically extract domain ontologies. Design/methodology/approach – The work here presents a Service Ontology Learning Framework (SOLF), the core aspect of which extracts Structured Interpretation Patterns (SIP). These patterns are used to automate the acquisition (from production domain specific Web Services) of ontological concepts and the relations between those concepts. Findings – A Semantic Web of accessible and re‐usable software services is able to support the increasingly dynamic and time‐limited development process. This is premised on the efficient and effective creation of supporting domain ontology. Research limitations/implications – Though WSDL documents provide important application level service description, they alone are not sufficient for OL however, as: they typically provide technical descriptions only; and in many cases, Web services use XSD files to provide data type definitions. The need to include (and combine) other Web service resources in the OL process is therefore an important one. Practical implications – Web service domain ontologies are the general means by which semantics are added to Web services; typically used as a common domain model and referenced by annotated or externally described Web artefacts (e.g. Web services). The development and deployment of Semantic Web services by enterprises and the wider business community has the potential to radically improve planned and ad‐hoc service re‐use. The reality is slower however, in good part because the development of an appropriate ontology is an expensive, error prone and labor intensive task. The proposed SOLF framework is aimed to overcome this problem by contributing a framework and a tool that can be used to build web service domain ontologies automatically. Originality/value – The output of the SOLF process is an automatically generated OWL domain ontology, a basis from which a future Semantic Web Services can be delivered using existing Web services. It can be seen that the ontology created moves beyond basic taxonomy – extracting and relating concepts at a number of levels. More importantly, the approach provides integrated knowledge (represented by the individual WSDL documents) from a number of domain experts across a group of banks.
International Journal of Web Information Systems – Emerald Publishing
Published: Aug 23, 2013
Keywords: Ontology learning; Semantic Web Services (SWS); Information extraction; Knowledge extraction; Web semantics architectures; Applications and standards; Web search and information extraction; Advanced web applications; Web data integration; Semantics; Knowledge management