Tuesday, 15 January 2002: 11:00 AM
Automated scanning of observational datasets for the generation of formal metadata
We have constructed the foundation for a framework of tools that can scan observational data to produce formal Federal Geographic Data Committee (FGDC) metadata. The framework provides a context in which a user can select various observational data sets and associate external metadata (such as contact information) with the selected datasets. The tools can then be invoked to extract metadata stored in these observational data sets and combine the extracted and external metadata to produce FGDC metadata for each of the observational data sets in a semi-automated fashion. The first implementation leveraging this framework extracts metadata from datasets using the EPIC netCDF conventions and places this information, along with other metadata, into the central metadata store (managed by a MySQL database). From there, the metadata can be extracted and converted into various formats, including formal FGDC format, XML-FGDC format and NOAA-FGDC format. The toolkit leverages state-of-the-art technologies including Java, XML, Extensible Sylesheet Language Transformations (XSLT) and the MySQL database system. Additional efforts are ongoing to increase the richness and robustness of the resulting metadata and to incorporate components to extract metadata from other observational dataset formats.