Towards an Active Directory of Geospatial Web Services

Spatial Data Infrastructures often maintain directories of public geospatial Web Services built from Web services listed in their registries. However, registry data may be stale and incomplete because their maintenance depends on service owners. A solution is to back the registry with an active directory, or index, that lists Web services found by a Web crawler focused on geospatial Web services. The problem of the identification of the geospatial nature of a Web service appears when such crawler finds a service described in the Web Services Description Language (WSDL). This paper presents two components required to build an active directory: a crawler focused on Open Geospatial Consortium and WSDL-based Web services, and an unsupervised classifier of geospatial Web services described in WSDL.