CATALOG SERVICE FOR THE WEB, DEVELOPMENT OF INTEROPERABLE SERVICE FOR DISCOVERING GEOSCIENCE RESOURCES

Formal metadata was invented to allow managing and searching for information resources based on domain and content specific criteria, to more efficiently locate and evaluate scientific or technical content that does not index effectively using automated text and http-link-based algorithms. There are compelling business reasons for standardized metadata content, format, and services: Improve efficiency searching for data and information; avoid use of incomplete, inaccurate, or superseded versions of content; avoid duplication of effort in searching multiple catalogs or developing software to interact with different catalogs.

To achieve interoperable catalog services, a community of practice must agree on a metadata content model, vocabularies to use as keywords to index content, and the protocol for searching catalogs and obtaining metadata. The ISO 19115 and 19119 specifications are becoming widely accepted as content models for metadata, but for application in a particular community, more specific profiles of these specifications are necessary. The U.S. Geoscience Information Network (USGIN) project a collaboration with State Geological Surveys, U.S. Geological Survey, an Energy Industry Consortium, and other academic and industry partners in the geoscience community to develop metadata profile for indexing geoscience resources. Various keyword thesauri are in use for indexing geoscience resources (e.g. GeoRef), but these have mostly been used in traditional library type systems, and must be adapted to operation in a distributed system linked by web services. Considerable work is also necessary to harmonize the various vocabularies and to develop standard encodings, service interfaces and semantic relationships. The USGIN project is implementing catalog services using the Open GeoSpatial Consortium (OGC) Catalog Service for the Web (CSW), testing free, open-source implementations by the Deegree and GeoNetwork projects. Currently, variations in interpretation or ambiguity in the specifications interfere with actual catalog interoperability, but problems are being identified, and the project will document ‘best practices’ to utilize the existing software to link multiple clients and servers.