> On Wed, Jul 27, 2011 at 8:59 AM, Jonathan Rochkind <rochkind_at_jhu.edu>
> wrote:
>> On 7/26/2011 8:50 PM, Karen Coyle wrote:
>>>
>>> I wouldn't say "much more" but at least a step in the right
>>> direction. We
>>> present dates to users as searches (from __ to ___) or as facets,
>>> and they
>>> have every logical right to think that this is the date of the
>>> WORK, not the
>>> date of a particular printing or re-publication.
>>
>> It is not as obvious to me as it is to you that this is always what
>> users
>> want or expect. I think sometimes they really DO mean to be
>> searching for
>> publications that were published on a certain date. (or sorting by
>> publication date).
>>
>> But clearly other times they want/expect/need searching by date of
>> the work,
>> not a particular republication.
>>
>> I guess we'd need some research to know how often they want one vs
>> the
>> other, in what contexts and circumstances.
>>
>> Or maybe not, and we just need to try and provide for both. But,
>> contrary
>> to some user expectations, the system can't read their mind and
>> figure out
>> which they meant, providing for both would need to be an additional
>> interface element of some kind where they chose what they wanted. I
>> suppose
>> we'd want the research to know what they wanted (in a given library
>> or user
>> community) most often, to make that the default, and the other
>> available
>> perhaps in some advanced options panel.
>>
>> But, yeah, either way it requires actually having the data in a
>> reliable and
>> useable way, which we don't really now.
>>
I agree that people do sometimes want to search by publication date,
but most of the time when they sort by pub date for nonfiction, I
suspect they're using it as a proxy for work dates.
For moving images, the main use case I could come up with for
publication date is a user who wants to see what's recently come out on
DVD or Blu-ray in some category (comedies of a certain era, very early
film, movies from a certain country, etc.) This is a scenario that is
probably more applicable to a content form that often undergoes shifts
in technological platform. Older movies that were not previously
available on DVD or Blu-ray continue to come out over time and a sort or
limit by publication date would be helpful for that.
For nonfiction videos, more often the date of publication is deceptive.
In my old job, I once had to rush catalog some DVDs on evolution that
were used in classes. The professor had been using VHS and wanted the
convenience of the DVDs. These were cataloged with a publication date of
2005, but the content was from 1981 and looked it.
Expression date is also likely to be useful in some cases (and perhaps
more than publication date if you think about it). The other date that
patrons are interested in is the date that the library acquired
something, particularly for new arrivals.
The OLAC FRBR-inspired demo search interface for moving images
(http://blazing-sunset-24.heroku.com/) does try to present both work and
publication dates. The date of the work is at the top in the work facet
section and is labeled "dates." One thing that became clear very quickly
is that we need to label this better, perhaps "original date." The
publication date is down the right side with
expression/manifestation/item limiters. We tried to use the layout to
clearly distinguish the work facets describing the original from the
facets for the other group 1 entities. I have mixed feelings about how
it worked in practice for users, but I think it was helpful to
demonstrate the concept.
Kelley
PS If anyone is interested in more info on the OLAC demo, Bill Kules,
Chris Fitzpatrick and I had a short paper at JCDL which I put up at
http://pages.uoregon.edu/kelleym/publications/JCDL_OLAC_FRBR_prototype.pdf
Received on Sun Aug 14 2011 - 20:50:42 EDT