Morphing towards interoperable catalogues

The World Wide Web (WWW) provides geo-spatial data custodians with an environment in which to advertise their datasets via on-line catalogue systems. The geo-spatial community is diverse and its members are spread around the world. The user's task of searching for and locating datasets of interest, would be greatly simplified if a single point of entry could search for datasets from multiple, independent catalogue systems-an interoperable catalogue system. Advances in middleware for distributed systems now make it straightforward to invoke remote catalogues. However semantic interoperability is still a challenge due to different catalogue systems employing different metadata standards. A metadata standard is a collection of attributes used to describe datasets. This paper employs ontologies as a framework to classify semantic mismatches between metadata standards before resolving them. Using ontologies provides a deeper understanding of the semantic mismatches. In addition, both an architecture for an interoperable catalogue system and the implementation of mediators, the component which resolves semantic mismatches, are described. The implementation uses emerging Web technologies such as XML and XSLT. To demonstrate the appropriateness of the approach, this paper shows how semantic mismatches are resolved between two well-established and widely used metadata standards; the ANZLIC (Australia and New Zealand Land information Council) metadata standard and the DIF (Directory Interchange Format) metadata standard.