[OAI-implementers] XML Schema for OAI compliance with NSDL harvesting

ePrints Support support@eprints.org
Fri, 18 Jan 2002 11:06:58 +0000

Sorry for replying to old message - I'm catching up on my mail.

On Thu, Dec 13, 2001 at 01:58:03PM -0400, John Perkins wrote:
> resources it would work fine. What we gained by many communities 
> adopting DC for discovery would far offset the lack of depth and 
> breadth at the discovery level.  We also concluded that using DC 
> qualifed was an additional level of complexity that didn't buy us any 
> more at the discovery level and so we suggested museums not use it.

I feel strongly that we should keep simple dc as required in OAI. Some 
archives really care about supporting OAI, but many more have no objection
if it's not too hard. DC is cheap and cheerful and at the very least better
than a google search (I would hope). It will promote broad resource 

Anyone who is willing to implement a more complex schema should be able
to implement unqualified dc easily "in passing".

The problem for this very loose format is for people trying to mechanically
process the OAI data. For example:
- identifying and merging records from different sources which describe 
the same item.  
- citation linking work. For example people wanting to build systems which
trawl huge piles of OAI records looking for the metadata record which best
matches a citation in a paper. 

Lack of semantics make this tasks harder. I know that some people on the
list work on this kinda stuff (Tim Brody), if I've missed the point, let me

If I were working on this I would want clearer semantics, but still having
all archives using the same semantics.

Maybe OAI could define a better but optional qualified dc standard which 
is a bit more work to implement, like level 2 DOM. Maybe "oai_lv2_dc" which
could make it clear that the date is the date of creation of the item
refered to by the metadata, not the record of it. And clearly define how
dates and peoples names should be formated.

This would require no change to the current protocol, just OAI recommending
a oai_lv2_dc spec as "best-er" practice. People would still have to do
oai_dc too. (which would be pretty damn easy).

The people wanting to use OAI for mechanical processing rather than resource
discovery would have to discuss the actual requirements.


 Christopher Gutteridge                   support@eprints.org 
 ePrints Technical Support                +44 23 8059 4833