Metadata Provenance
Revision as of 22:59, 1 September 2010 by Nathan Yergler (talk | contribs)
| Contact | Contact::Nathan Yergler |
|---|---|
| Project | ,|project_name|Project Driver::project_name}} |
| Status | Status::Complete |
The initial version of DiscoverEd does not include provenance support. Provenance means tracking the source of resource metadata. Due to this limitation, DiscoverEd has limited ability to filter by curator. While you can filter for resources with a specific curator, the remaining search terms are not limited to metadata provided by that curator. This is a significant shortcoming for resources with multiple curators.
Requirements
- The provenance of metadata discovered through RSS, Atom, and OAI-PMH is stored in the RDF Store.
- Metadata extracted from structured data is stored with provenance reflecting the page it was extracted from.
- Users can filter a query to exclude a curator, and metadata provided by that curator is not considered for other query terms. For example, "
excludecurator:http://example.org subject:biology cells" would return results containing the term "cells", with the subject tag "biology" provided by a curator other than http://example.org.
Status
Provenance support was initially added with table prefixes, and later refactored to use Named Graphs for Jena. Provenance support has been landed in next, and is running on Labs.
As of 30 August 2010, provenance support has landed in master.
