This paper documents the use of unqualified Dublin Core (uDC) elements in the metadata harvested from the repositories participating in the CARL harvester, and identifies patterns in the use
of that metadata. It also compares these findings with a similar study, and identifies areas for further research.
