Semantic Interoperation for GeoScience Models
View/ Open
paper1.pdf (444.4Kb)
paper2.pdf (698.9Kb)
Supporting document.pdf (593.4Kb)
appendices.pdf (101.6Kb)
Date
27/11/2007Item status
Restricted AccessAuthor
Qiang, Yi
Metadata
Abstract
In recent years, integrated models involving components from different disciplines have become the developing trend of GeoScience modelling. However, a common lack of semantic representation produces hurdles for model sharing, reusing and interoperation. SOA and Web Service provide a more interoperable architecture for GeoScience Modelling with distributed components, however, is still limited to semantic interoperability. This paper theoretically discussed the motivation of using Semantic Web to interoperating GeoScience models, and reviewed the development of Semantic Web and its application in GeoScience modelling, and then discussed the semantics involved in SOA-based GeoScience models, finally conclude this paper and propose future work. With the growth of modelling in the GeoSciences, many difficulties have occurred in integrating model components and retrieving geospatial information because of their diversity. Web Services and SOA provide an advanced architecture for distributed, dynamic, flexible, and re-configurable service systems across the Internet, which also shows great potential in service-enabled GI systems and GeoScience models. Due to the inherent deficiency of semantic interoperability with XML-based Web Service specification, they are still limited in service discovery, invocation and composition. These deficiencies are effectively solved by Semantic Web Services which describe Web Service in a computer-interpretable structure. After reviewing relevant technologies in Web Services and Semantic Web Services, this paper utilizes two specific models to assess the feasibility of OWL-S (one of the mainstream Semantic Web Service standards) in interoperating GeoScience model components, and discusses some possible semantic representations of geospatial information. This research reveals problems in the semantic interoperation of model components, and identifies future work needed to solve these problems.