Authors:
Aleksandar Stanimirović, Miloš Bogdanović, Leonid Stoimenov
Keywords:
ontology, interoperability, data semantics
Publication:
In Proceedings of 12th AGILE Conference on GIScience
Issue date:
2009
Abstract:
Nowadays, information systems (IS) use the Internet/Intranet as their natural operating environment. Internet/Intranet environment has introduced additional opportunities in the development and operation of information systems and sources of information they use. Information systems take advantage of multiple heterogeneous sources of information. Using information from different sources includes the ability of these applications to integrate information. Effective information integration from heterogeneous sources relies upon semantics of data that is to be integrated.
In order to enable information integration, it is necessary to make information both machine readable and machine understandable. Ontology and semantic languages give us an opportunity to define a meaning of the information and machine-processable semantics of information sources. Ontologies are used by different ISs that participate in communication in order to solve the meaning of the available data. Ontology development and usage is also present in the area of a Geo-Information System (GIS). Geo-information domain has characteristics that further complicate the semantic description of geospatial information necessary to achieve GI systems interoperability.
Geospatial data semantics deals with representations of geographical world as interpreted by human users. Representation and reasoning on the meaning of geospatial data are critical for the development of interoperable geospatial data and software, geographical information retrieval, and automated spatial reasoning. This unavoidably includes development of spatial ontologies. Well designed spatial ontologies could supply both stand-alone applications and search engines with localization information and thus refine queries and enrich the retrieved information. However, as a result of complex, heterogeneous, multimodal and multimedia representations of spatial data, it is very difficult to capture and maintain semantic knowledge of geographical data. At the moment, geographical information systems either impose a simple semantic structure or do not address the issue of semantics at all, but simply offer textual metadata description, leaving the burden of meaning construction to the user. Such solutions are inadequate in the current stage of geographical information technologies where massive exchange of data from heterogeneous sources must be supported.
Ontology usage intensification in the IS development could be achieved if we treat ontologies not only as a means of semantic mediation of the existing ISs, but as a starting model for the IS development process. This approach, used for information integration, demands ontology-to-data source mapping. The process of ontology-to-data source mapping is the first step in making IS data sources integration possible. Ontology-to-data source mapping will add semantic description layer to each of ISs involved in the process.
URL:
http://www.ikg.uni-hannover.de/agile/fileadmin/agile/paper/60.pdf