| Summary: | An ontology provides an explicit description of the concepts and relationships within a particular domain. They are used within computer science with the aim of enabling more effective data integration between heterogeneous datasets. The principal goal of ontologies, that of fully automated knowledge interoperability between computers, has y.et to be achieved. Within ontological research, there is no standard approach for ontology development, and one of the main aspects - reuse of existing ontologies, is never fully described or evaluated. At present, ontologies are the forte ofthe logician due to the complex nature ofthe formal logic they are represented in. The increasing variety of domains using ontologies requires the domain expert to have a central role in the ontology development process, something that is rarely recognised within the literature. This research explores two avenues of potential benefit to the ontology development process: the reuse of existing ontologies; and the role ofthe domain expert. This required two new methodologies to be established for building conceptual. ontologies, either from first principles or through recycling components of an existing ontology. Conceptual ontologies provide a concise and less ambiguous representation which is an intermediary between natural language and the complex languages used for formal ontologies. These were used throughout this research as a means of enabling the domain expert to construct ontologies that they are able to understand and verify. A case study based on environmental statutory bodies was used to evaluate whether these methods could successfully be used by a domain expert to construct conceptual ontologies. The results showed that the domain expert was fully capable of building conceptual ontologies and can therefore be placed at the forefront of the ontology development process. The conceptual ontologies can then be transformed into a formal representation by an ontology engineer as required. The recycling aspects were also evalu'ated and it was found that improved efficiency in development will only be achieved when there is there is a good degree of correspondence between the concepts under consideration. There are, however, less tangi.ble benefits associated with recycling since there is more consistency between.the new and recycled ontology, which will improve interoperability between datasets based on these ontologies.
|
|---|
