Knowledge representation and semantic interoperability in marine information systems

Several aspects of marine environmental information systems motivate this research. Datasets for marine GIS (Geographical Information Systems) that are resource-intensive to collect and maintain have to be shared and exchanged among the research community in order to carry out a range of spatiotemporal analyses. The technological infrastructure to enable this to occur is available in the form of in-house databases, external data sets, the World Wide Web and the Internet. However, the use of common terminology is less widespread, thus making it difficult to ensure that information resources can be easily integrated. International research initiatives such as the EUNIS habitat classification seek to provide a domain-relevant perspective on the concepts and terminology used by researchers to refer to marine spaces. The notion of expert, semantic domain knowledge is introduced to enable disparate marine data sources to be specified using a common terminology. This knowledge can also be used to provide intelligent user support by enabling aggregation of information in the temporal, spatial and scientific dimensions at levels appropriate to research problem being studied. Key words: habitat classification, marine information system, ontologies, semantic knowledge