Ross wrote:
"It's not a valid argument to say that this would work in a subject
browse, because our stats overwhelmingly indicate our users are only
using the keyword anywhere search and nothing more. Theory behind
subject headings be damned when the actual user behavior rears its
head."
But I think we should consider that Endeca or Scriblio systems might
change that. By leading people to controlled vocabulary via particular
titles of interest, they might increase people's interest in subject
searches.
Also, if you're going to talk about stats, at least make an effort to
determine what populations generate which stats. Yes, I assume that
freshmen use keyword searches overwhelmingly. But by the time they're
seniors, have they learned to use subject searches? I realize you may
have done that already, and I'm open to correction on the point.
My opinions are influenced by Thomas Mann's experience of what he's
found helpful to researchers over 30+ years as a reference librarian, at
least half of that time in an online context. Of course statistics have
value, but I think they can be overvalued, too. If I knew from personal
experience that subject searches were helpful to my patrons, would I
give them up just because there's some statistical study showing the
"average" patron doesn't use them?
--Ted Gemberling
-----Original Message-----
From: Next generation catalogs for libraries
[mailto:NGC4LIB_at_LISTSERV.ND.EDU] On Behalf Of Ross Singer
Sent: Wednesday, May 09, 2007 9:08 PM
To: NGC4LIB_at_LISTSERV.ND.EDU
Subject: Re: [NGC4LIB] Yes but
Out of curiosity, what sort of interface handles this ideally?
I feel Endeca's (et al.) keyword search is being compared to something
that's not a keyword search. Now, if your catalog is set up like my
catalog and your default search is keyword anywhere, how are your
results better than what Scriblio or Endeca return?
It's not a valid argument to say that this would work in a subject
browse, because our stats overwhelmingly indicate our users are only
using the keyword anywhere search and nothing more. Theory behind
subject headings be damned when the actual user behavior rears its
head.
-Ross.
On 5/9/07, Ted P Gemberling <tgemberl_at_uab.edu> wrote:
> Yes, Jonathan, I'm sure FAST is a serious attempt at faceting. (I
stand
> corrected on my earlier claim that Casey's subject display isn't
> faceted.) And one could say that as long as "philosophy" and
"influence"
> are present somewhere in the 650 fields, it doesn't matter whether
> they're coordinated with Plato and Shelley. But I do think there's
also
> a kind of "aesthetic" element that comes in here: there's something
> about seeing "influence" alone on a screen that bothers even the
> developers of FAST--I can testify to that because I was at their
meeting
> in 2005 when they said so. Somehow Philosophy alone looks better than
> Influence alone, I suppose because Philosophy seems coherent as a
> subject on its own.
>
> Also, if you decompose those subheadings, including Philosophy,
subject
> browse screens become pretty useless. People are going to be less
> interested in clicking on Philosophy on the browse screen if they know
> that in many cases the titles are not really works on philosophy but
on
> some person's philosophy. The case with Influence would be even worse.
>
> I don't mind it much that Endeca and Casey's catalogs decompose the
> headings in the "faceted" sidebar displays. That's probably necessary.
> But part of the reason it doesn't bother me is that (at least) Endeca
> doesn't abolish alphabetical subject displays. And I think Casey's
> doesn't, either. Both kinds of searches are options.
> --Ted Gemberling
>
> -----Original Message-----
> From: Next generation catalogs for libraries
> [mailto:NGC4LIB_at_LISTSERV.ND.EDU] On Behalf Of Jonathan Rochkind
> Sent: Tuesday, May 08, 2007 3:39 PM
> To: NGC4LIB_at_LISTSERV.ND.EDU
> Subject: Re: [NGC4LIB] Yes but
>
> I'm also interested in someone exploring what makes a controlled
> vocabulary suited for the kind of facetted exploration we are talking
> about. LCSH was not created with this in mind, it was created thinking
> that users would 'interface' to it primarily through an alphabetic
> listing of all existing pre-formed headings. Many of it's features can
> be only understood with that context in mind (the "--" subdivisions
and
> the way they work, for instance. I'd like to figure out a way to put
> the _hiearchical_ relationships in LCSH into an easy to use browse
> interface, but the fact that there are at least two, if not three,
> seperate hiearchical structures within LCSH, doesn't make this easy).
>
> What features should a controlled vocabulary have to make it suitable
> for a facet-style exploratory use? What are the different contexts a
> controlled vocabulary can be used in? What features are helpful for
> each? Do some come at the expense of others? How can you effectively
use
> LCSH in the kind of easy to use exploratory interface Casey talks
about?
> How would you design a controlled vocabulary from scratch for that
> purpose? What reasonably efficient interventions can be made in LCSH
to
> improve it's use in easy to use interfaces? (And what sorts of easy to
> use interfaces are possible?).
>
> These are the questions we need to explore, with both intellectual
> analysis, experimental projects, and empirical user-centered research.
>
> The FAST project is to some extent an attempt to make LCSH more
amenable
> to this type of interface, but to my mind it's only the barest
beginning
> of what's possible. It's not backed by much empirical research. It
loses
> some information from LCSH that I'm not convinced you _need_ to lose
(is
> it possible to keep this information intact and even present it in an
> easy to use manner?). And it's generally presented as a project to
"make
> subject cataloging easier for paraprofessionals", rather than "make
> subject controlled vocabulary work better in easier to use
interfaces."
> It's the latter I'm interested in, not the former. I think a lot more
> than FAST, and not neccesarily along the lines of FAST either, is
> possible.
>
> Jonathan
>
> Casey Bisson wrote:
> > Ted,
> >
> > This is an excellent example.
> >
> > I often ask people if they know what "bagged products" are, and the
> > usual answer is "huh?" Then I offer this picture (link below) and
> > watch as people immediately understand the term.
> >
> > http://maisonbisson.com/blog/post/11538/
> >
> > I'm an advocate for kind of controlled vocabularies you describe
here,
> > but I've also seen how we can represent them in our systems in ways
> > that help the user make better sense of them.
> >
> > Example: I often see "sociology of education" appear in our search
> > stats, while the correct LCSH is "educational sociology." Clearly
> > there's a huge number of users at my library that don't know the
LCSH,
> > but they still need good results. My solution (and it's old hat by
> > now) was to display the aggregate subjects as a facet.
> >
> > http://plymouth.edu/library/opac/search/sociology+of+education
> >
> > And using your examples, the subject facets again reveal some very
> > useful information:
> >
> > http://plymouth.edu/library/opac/search/eskimo
> > http://plymouth.edu/library/opac/search/inuit
> >
> > The challenge I'm trying to meet is to provide sophisticated results
> > without increased complexity. The subject facets reveal what the
> > catalog knows (based on what librarians have acquired and the
metadata
> > they have) about the keywords the user searched. We know from
previous
> > studies that users modify their searches based on the results
> > returned, and I've seen lightbulbs appear in users as the explore
the
> > facets.
> >
> > The result is that a user who didn't know the LCSH before starting a
> > search learns it quickly.
> >
> > That is, sophisticated tools can make complex research easy.
> >
> > Now one of the things I'd like to see is tooltips for the LCSH
facets
> > that offer a deeper explanation of what they are (and are not).
> >
> > Notes:
> > 1: the code serving the above links is over a year old and is
> > embarrassing, but it's got the largest collection of relevant items.
> > For a more interesting and up to date example of Scriblio (was
WPopac)
> > see http://beyondbrownpaper.plymouth.edu/browse/ .
> > 2: my library's collection doesn't come close to serving the needs
of
> > somebody researching "Judaism and the difference between its
concepts
> > of Messiahship and those of Christianity," the first example in your
> > original message.
> >
> > --Casey
> >
> >
> > On May 8, 2007, at 3:15 PM, Ted P Gemberling wrote:
> >
> >> Here's another example that shows the important role of librarians
as
> >> information "experts." A lot of people today are under the
impression
> >> that "Inuit" and "Eskimo" are equivalent terms. Generally Inuit is
> >> considered more appropriate to use. NLM's Medical Subject Headings
> >> accept that equivalence and establish Inuit as the term. But if you
> look
> >> at the LCSH hierarchy, you find that Eskimo is actually a broader
> term
> >> than Inuit. Here's the scope note for Inuit:
> >>
> >> "Here are entered works limited to the indigenous Arctic peoples of
> >> Greenland, Canada, and northern Alaska. Works discussing
collectively
> >> the Inuit peoples and the related Eskimo peoples of southern and
> western
> >> Alaska and adjacent regions of Siberia, or works for which the
> >> individual group cannot be identified, are entered under ǂa
Eskimos."
> >>
> >> Probably 70-80% of all Eskimos in the world are Inuits, but having
> spent
> >> one summer in Western Alaska, I'm aware there is another 20-30% who
> are
> >> Yupiks. The only term we have for both groups is Eskimos. This
shows
> the
> >> close collaboration LCSH subject specialists have with people with
> >> knowledge of subject areas. Just looking at the LCSH syndetic
> structure
> >> is informative for a researcher. Keywords cannot provide that
> >> information without a lot more work on her part.
> >
> >
> > Casey Bisson
> > __________________________________________
> >
> > Information Architect
> > Plymouth State University
> > Plymouth, New Hampshire
> > http://oz.plymouth.edu/~cbisson/
> > ph: 603-535-2256
> >
>
> --
> Jonathan Rochkind
> Sr. Programmer/Analyst
> The Sheridan Libraries
> Johns Hopkins University
> 410.516.8886
> rochkind (at) jhu.edu
>
Received on Thu May 10 2007 - 14:40:32 EDT