Hi Karen,
Thanks for taking time to think about my inquiry and responding.
Sadly, Harvard's Hollis still buries individual results in little
pockets that must each be opened and closed separately.
This is especially bad for subject headings displays. Subject
subdivisions now effectively separate a record from others in the same
main subject by hiding them in separate little pockets. Should we begin
repeating the main heading in a separate subject field in order to
visually bring things back together? A better display would be the
preferred solution.
Our previous system here featured a "merged headings index" that
enabled the desirable display features. I've been snooping around the
major libraries' catalogs and been quite disappointed with their browse
results displays. Since these are typically Oracle-based, is it a
problem inherent in Oracle databases that they would be incompatible
with a merged headings index? Isn't the FRBR phenomenon demonstrative of
the importance of relationships in bibliographic service? Unit records
could (in my imagination at least) both be stored whole and separately
broken up as needed for more useful navigation and display.
Economics will probably rear its ugly head again, but to me
nothing's too good for my library patron community. Neither keyword nor
browsing offers a complete finding package for searchers. Maybe if your
library's searchers aren't too demanding and developing browse is
expensive, then you can let them eat keyword. Mine also include serious
scholarly researchers whose needs are more exacting. Again, I beg for
better browse and keyword searching, not just one or the other.
Thanks and have a great day,
Jimmie
-----Original Message-----
From: Next generation catalogs for libraries
[mailto:NGC4LIB_at_LISTSERV.ND.EDU] On Behalf Of Karen Coyle
Sent: Monday, July 30, 2007 3:56 PM
To: NGC4LIB_at_LISTSERV.ND.EDU
Subject: Re: [NGC4LIB] Better browse
Jimmie Lundgren wrote:
It would be helpful after choosing one of these
> headers and opening it if upon returning to the results list each of
the
> titles were displayed indented below the corresponding specific
> headings. This enables much smoother and more efficient navigation for
> selecting useful items than hiding them in little pockets under the
> heading strings.
Harvard's Hollis catalog does something of this nature, although the
interface leaves me a bit cold. You do have to go to another screen to
see the entries under each author. Do an author search and then click on
the button "sorted index" to the right.
This requires some gyrations in the database design. I can't go further
with that without diagrams, but let me just say that a very different
design would be needed, and I'd love to hear design ideas. I think it
would require us to give up our "unit record" -- that is the monolithic
MARC-based record that gets stored in most databases (even if not in
true iso 2709 format). Creating a database that emphasizes relationships
(authorx wrote booky, authorx wrote bookz) rather than a bibliographic
record would make this kind of display more possible. Today, when you
search in a library catalog you are getting back a set of these rather
impenetrable unit records, and a fair amount of work would have to be
done to pull out the relationship data for display. This is why faceting
is looking so interesting because it pulls out and analyzes information
that usually stays hidden in the bibliographic records. I don't know if
similar software could create the display you desire.
kc
--
-----------------------------------
Karen Coyle / Digital Library Consultant
kcoyle@kcoyle.net http://www.kcoyle.net
ph.: 510-540-7596 skype: kcoylenet
fx.: 510-848-3913
mo.: 510-435-8234
------------------------------------
Received on Tue Jul 31 2007 - 12:49:30 EDT