Deriving an Australian marine ontology from existing ontological models: a practical evaluation.

Domain and purposive ontologies provide explicit and formal definitions of real-world concepts. They are indispensable components of any system whose objective is to support the exchange and semantic integration of data and information. The reuse of existing formalised ontologies is often encouraged to reduce development overheads and increase semantic interoperability between data service providers. This paper presents some selection criteria that are deemed important for choosing potentially reusable ontologies when dealing with marine science data. An Australian marine science case study involving the Integrated Marine Observing System (IMOS) was used to investigate ontological requirements. Climate Science Modelling Language (CSML) and Observation and Measurement (O&M) were the two main ontologies under consideration for re-use. By adapting an ontology construction methodology, for use instead as an evaluation tool and by using sample data from the IMOS project, CSML and O&M were evaluated for their suitability to support the exchange of data within the IMOS information infrastructure. As a result of the evaluation exercise a merged CSML and O&M ontological model was deemed preferable to using either ontology on its own.