National Spatial Data Infrastructure Partnership Programs: Rethinking the Focus (2001)
Chapter: Don't Duck Metadata
core team for all of the projects tends to be small, and the temporary nature of the funding often leads to an unstable working environment. The departure of a key player can severely impede the success of a project and momentum can quickly disappear.
Don’t Duck Metadata
In 1999 the FGDC sponsored 95 projects to promote the creation and use of metadata in support of geospatial data sharing. This program was designed to encourage the adoption of consistent policies for metadata, and to counter the notion that metadata are expensive to create and have limited benefits. Grants of approximately $18,000 were given to 42 states to stimulate partnerships that would promote the development of metadata.
Metadata play a critical role in the NSDI. They facilitate the sharing of data, particularly between partners who are not in direct contact with one another; it is necessary to document the contents of datasets; to provide sufficient detail to allow computing systems to open and access them; and to document data quality. In effect, these metadata components allow potential users to assess the fitness of datasets for their own use, and to minimize the problems associated with importing data from another system.
Such sharing of data is central to the NSDI goals of reducing duplication of effort, improving data quality, and improving data access. Unfortunately the benefits and costs of metadata creation accrue in ways that do not necessarily promote these goals. Most of the costs of metadata creation accrue to the custodians and creators of data, while most of the benefits accrue to users, often in other organizations. As a result, data providers tend to “duck” metadata or to assign them a low priority. The FGDC believes that one solution to this difficulty is to bring users and creators into a single partnership that can reassign or aggregate costs and benefits in ways that are more satisfactory to all the partners.
The committee considers that smaller grants (e.g., the average award of $18,000 in 1999, and $22,200 in 2000) appear to be inadequate to meet the program’s objectives. Moreover, the decision to fund almost all applicants (95 of 108 in 1999; 31 of 32 in 2000) may prove to be
