USGIN ISO metadata profile

Abstract

The USGIN project has drafted and is using a specification for use of
ISO 19115/19/39 metadata, recommendations for simple metadata content,
and a proposal for a URI scheme to identify resources using resolvable
http URI's(see http://lab.usgin.org/usgin-profiles). The principal
target use case is a catalog in which resources can be registered and
described by data providers for discovery by users. We are currently
using the ESRI Geoportal (Open Source), with configuration files for the
USGIN profile. The metadata offered by the catalog must provide
sufficient content to guide search engines to locate requested
resources, to describe the resource content, provenance, and quality so
users can determine if the resource will serve for intended usage, and
finally to enable human users and sofware clients to obtain or access
the resource. In order to achieve an operational federated catalog
system, provisions in the ISO specification must be restricted and usage
clarified to reduce the heterogeneity of 'standard' metadata and service
implementations such that a single client can search against different
catalogs, and the metadata returned by catalogs can be parsed reliably
to locate required information. Usage of the complex ISO 19139 XML
schema allows for a great deal of structured metadata content, but the
heterogenity in approaches to content encoding has hampered development
of sophisticated client software that can take advantage of the rich
metadata; the lack of such clients in turn reduces motivation for
metadata producers to produce content-rich metadata. If the only
significant use of the detailed, structured metadata is to format into
text for people to read, then the detailed information could be put in
free text elements and be just as useful. In order for complex metadata
encoding and content to be useful, there must be clear and unambiguous
conventions on the encoding that are utilized by the community that
wishes to take advantage of advanced metadata content. The use cases for
the detailed content must be well understood, and the degree of metadata
complexity should be determined by requirements for those use cases. The
ISO standard provides sufficient flexibility that relatively simple
metadata records can be created that will serve for text-indexed
search/discovery, resource evaluation by a user reading text content
from the metadata, and access to the resource via http, ftp, or
well-known service protocols (e.g. Thredds; OGC WMS, WFS, WCS).