I can give some more information about VIAF and WorldCat Identities,
since I'm involved in both here at OCLC Research.
The Virtual International Authority File (VIAF) should facilitate the
identification of controlled forms of names and linking between national
authority files. So, given the form of the name established in Germany
at the DNB it would be possible to find the NACO equivalent. Currently
VIAF is a joint project involving LC, DNB and OCLC. We hope to add the
French soon. Nothing is public yet, but we're getting close to having
services (initially SRU, but some sort of OpenURL is also a possibility)
available. Currently there are about 600,000 links between the LC and
DNB forms of names, covering most of the more common names in both
files.
For Identities, (http://orlabs.oclc.org/Identities/) we're spending most
of our time 'productionizing' it. That means fixing bugs, getting it on
a regular production server here at OCLC, and using standard identifiers
in the URI's. It is available now though SRU, although there will be
some changes over the next month or two as we move it, and we expect an
OpenURL API to be available. Identities concentrates on names found in
WorldCat, but the records (everything is in XML) will have the cross
references we find in VIAF, as well as references to things like
Wikipedia that aren't participating in VIAF.
How easy it will be to incorporate these into local services is yet to
be seen, but we would like that to happen and are open to suggestions.
-----Original Message-----
From: Next generation catalogs for libraries
[mailto:NGC4LIB_at_LISTSERV.ND.EDU] On Behalf Of Bernhard Eversberg
Sent: Thursday, August 23, 2007 7:54 AM
To: NGC4LIB_at_LISTSERV.ND.EDU
Subject: Re: [NGC4LIB] content beyond books
Weinheimer Jim schrieb:
>
>> And an important one too. It gave rise to the VIAF project currently
>> under research at OCLC.
>> Ultimately, one can imagine an "Authority Google" of sorts, a
>> database somewhere in the background that can be consulted
>> automatically and transparently by catalog systems (or search engines
>> too) every time a personal name is requested. It would return (in the
>> positive case) an appropriate ID with which the catalog in question
>> can do the actual search.
>> With this concept in place, and authority files from the world over
>> merged into it eventually, libraries could well stick with their
>> national conventions for name headings instead of kindly asking their
>> patrons to find and enter enter LC forms of names.
>
> I hope this project is successful. From my understanding of it, it
will link the two by providing a link from heading to heading. Will
German users get the advantage of searching LC cross-references and
English users be able to use the German cross-references?
>
Well, it seems that "WorldCat Identities" emerges from this project,
although nothing much can presently be found out about it. But I'm sure
someone from Frankfurt or Dublin (Ohio) is listening in and ready to
enlighten us.
One easily fancies that they publish an API enabling everbody to
integrate it with their systems. (This is what would have happened were
they Google, anyway, but they will be lacking the
resources.)
But I may also be wrong and all the documentation has already been
rolled out and the API deployed - one cannot know everything.
B. Eversberg
--Thomas Hickey, Chief Scientist, OCLC
--614.764.6000
--mailto:hickey_at_oclc.org
--http://errol.oclc.org/laf/n82-54463.html
--http://purl.oclc.org/hickey/outgoing
Received on Fri Aug 24 2007 - 09:48:10 EDT