Also, figuring out if two marc records exported from two different
libraries represent the same 'thing' is non-trivial, due to the nature
of our data. OCLC provides a variety of services to make that easier,
including simply the fact that catalogers manually attach OCLC numbers
to their records (or manually download an OCLC record with an OCLC
number) -- an OCLC service.
Ross Singer wrote:
> On Tue, Mar 9, 2010 at 11:34 AM, Eric Lease Morgan <emorgan_at_nd.edu> wrote:
>
>> I beg to differ; it is trivial to dump and expose one's bibliographic records to the Web.
>>
>> Any ILS worth its weight in salt has an export function.
>>
>
> I think you may be surprised that this is not a universal truth.
> *Many* ILSes have an export function, some are premium features.
>
> I have no idea whether or not hosted ILSes (or which hosted ILSes)
> have this option (especially for free).
>
> And I'm still not sure I see where the value lies in all libraries
> exporting their data. Won't there be extremely high percentage of
> overlap for all but a small minority of libraries (which is where the
> really interesting data would be)?
>
>
> David brings up a valid point, as well. Taking the LC records dump in
> the Internet Archive as an example.
>
> 1) This is a snapshot from 2007
> 2) It's only book records
> 3) This is now out of date.
>
> So merely exporting our records isn't enough. How do we deal with the
> deltas? Does every library have to deal with the deltas?
>
> -Ross.
>
>
Received on Tue Mar 09 2010 - 13:06:07 EST