Distributed GIS and metadata : methods for the description of interoperable GIS components

An increasing number of GIS applications rely on computer networks and the World Wide Web for accessing their resources, being both datasets and software operations. Constructing applications in such a distributed GIS environment needs proper solutions for easy discovery of the necessary software and data components. Metadata is of primary importance for the identification of the connectivity between these components. In this paper we present a method to describe GIS components and their inter-relationships with XML and compare this method with the possibilities of the Resource Description Framework (RDF). A good description framework is flexible and at the same time rigid enough to enforce interoperability between operations and data. It has to fulfil the needs of different types of user as well as the requirements of the component providers. In this respect, XML has the advantage of being extensible, and both machine and human readable, but is open enough to trigger developers in writing their own, again inflexible, XMLbased description language. In our test environment XML proved to be a satisfactory description language that allowed a structured way of identifying the interoperability of a limited number of GIS components. However, the connection between these components was fairly controlled. RDF, while still XML-based, provides and prescribes more specific handles for resource descriptions with the help of Uniform Resource Identifiers. This allows for uniform discovery of globally unique resources (data and software). Such a concept is promising, but needs a clear understanding amongst the global user community of the descriptive elements within the GIS component space.