Multi-ontology web services discovery

• Main Author: Le, Duy Ngan
• Other Authors: Goh Eck Soong, Angela
• Format: Theses and Dissertations
• Language: English
• Published: 2008
• Subjects: DRNTU::Engineering::Computer science and engineering::Information systems::Information systems applications
• Online Access: https://hdl.handle.net/10356/13591
• Institution: Nanyang Technological University

The Semantic Web promotes machine understandability in the Web through the use of ontologies. An ontology is a specification of classes, properties, and their relations, from which new knowledge can be inferred. By enhancing traditional Web services with Semantic Web technology, Semantic Web services have emerged as a promising means of enhancing e-business and e-commerce operations. As a result, Semantic Web services have been rapidly developed. Though the numbers may not be as large as current Web pages and non-semantic Web services, the increasing use of Semantic Web service description languages such as OWL-S indicates that this number will increase further in future. This has led to a demand for a discovery mechanism. Researchers have developed discovery systems to discover Semantic Web services which satisfy the requirements of a requester. Current discovery systems locate Semantic Web services that are based on the same ontology but they do not support the discovery of Semantic Web services that use different ontologies. Thus, even if the Semantic Web service providers meet the requirements of a Semantic Web service requester, such a match may be missed. Unfortunately, in the real world, it is highly probable that requesters and providers utilize different ontologies to describe their services. Therefore, a discovery system that supports Web services using different ontologies is important. We have developed MOD, a Multi-Ontology web services Discovery system, which matches Semantic Web services described by the same as well as different ontologies. Techniques related to ontology and Web service matching such as ontology comparison and concept similarity measurement algorithm have been investigated. Ontology comparison determines if two ontologies are related. It is employed in measuring concept similarity. If it can be determined that two concepts, which originate from two different ontologies, actually belong to a common ontology, the accuracy of the measurement can be enhanced. Moreover, considering domain similarity has avoided mismatches and improved the matching of the overall concept similarity. In turn, concept similarity measurement is employed to measure Web service similarity which is the core of MOD. The proposed techniques are not only used in MOD but are also applicable to other ontology and Semantic Web service activities. MOD is superior to existing Web service discovery systems as it supports multiple ontologies and enhances the accuracy of Web service discovery.