Matching Schemas for Geographical Information Systems Using Semantic Information

Integration and interoperability is a basic requirement for geographic information systems (GIS). The web provides access to geographic data in several ways: on the one hand, web-based interactive GIS applications provide maps and routing information to end users; on the other hand, the data of some GIS can be accessed in a programmatic way using a web service. Thereby, the data is made available for other GIS applications. However, integrating data from various sources is a tedious task which requires the mapping of the involved schemas as a first step. Schema matching analyzes and identifies similarities of two schemas, but all approaches can be only semi-automatic as human intervention is required to verify the result of a schema matching algorithm. In this paper, we present an approach that improves the matching result of existing solutions by using semantic information provided by the context of the geographic application. This reduces the effort for manually correcting the results which has been validated in several application examples.