[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Ref-Links] Draft Document on DOI Infrastructure




Please excuse if you've been cross-listed on this posting. Thanks, Tony

---------------------- Forwarded by Tony Hammond/AP/LDN/HARCOURT on
14/12/2000 14:32 ---------------------------


Tony Hammond
13/12/2000 16:16

To:   discuss-doi@doi.org
cc:

Subject:  Draft Document on DOI Infrastructure


Attached is a (preliminary) draft for comment:

   "Towards an Implementation of DOI Application Profiles within the Handle
   System:
   A hierarchical architecture for DOI infrastructure"

It basically describes how DOI Application Profiles (was DOI Genres) can be
implemented using multiple resolution. Architecture highlights:

·    Every DOI (application and administrative) has one or more associated
application profiles.

·    All DOI handles are interlinked within the DOI handle application.

·    The DOI handle application is a self-contained implementation within
the Handle system and is
independent of other handle applications and intentions (eg handle type
semantics).

·    XAD - An eXtensible API for DOI is required. A preliminary is given.


Hope this might be of some interest and not too abstruse. There's a picture
inside and lots of "actionable" (virtual) DOIs. What more can I say?
RDF-compliant, even :-)

Disclaimers:
This is still an early draft and some of the language may be a little
uneven regarding DOI vs Digital Object, and Application vs Application
Profile. Also I've used a DOI genre now as shorthand for a DOI content type
(Journal Article, say) - this may still be misleading although seems (to
me) to be useful as the new concept of Application Profile has a better
tie-in with application-specific metadata. I've not distinguished too well
the difference between services for acquiring kernel and application
metadata in the CrossRef application - we're still working on that. I've
tried also to stike some kind of balance between specificity and genericity
in the DOI handle types although there's lots of scope for improvement.

Be glad to get any feedback.

Tony

(See attached file: DOI Infrastructure.doc)




=?iso-8859-1?Q?DOI_Infrastructure.doc?=