Well, of course there’s Dublin Core and its core elements. The ISO put out an update about it recently. Trying to get hold of that. Meanwhile I have looked over the 2007 report. Nothing terribly new from the last report. Just seems to be clarifications on terms. However, unqualified DC is what the OAI-PMH protocol is based on. Many large institutions have adopted this protocol – or variations on it with requests in GET/POST HTTP and responses in XML.
Although there are others – GLIS, TEI Headers, EAD, FGDC – (all good ones) – I really feel good about zeroing in on DC with perhaps differing profiles dependent on document type or subject area, for instance. The number of profiles should be kept to a minimum, accepted as written, with contingency for exceptions through a separate profile.
I think it’s good to keep in mind some differing categories of metadata when formulating the differing profiles – administrative, rights, descriptive, technical. I think also a controlled vocabulary should be decided upon for subject terms.
A few pieces I’ve been reading:
DC Metadata element set – ANSI/NISO Z39.85 2007
Guidelines for implementing DC in XML – Andy Powell, University of Bath, 2002
Metdata made simpler, Gail Hodge, NISO, 2001
