[Web4lib] Faceted navigation as metasearch

[Walker, David]

>> Anyone know a system that can do this?
 
Version 4.0 of Metalib [1], due out this spring, will include clustering using the Visimo clustering engine [2], which creates facets on the fly.
 
 
>> I think that pulling in all (or even some) 
>> of the retrieved records and creating the 
>> facets on the fly would be prohibitively slow.
 
I agree that pulling back all of the results would be prohibitive.  
 
Metalib pulls back the top 30 results from each database.  I haven't seen the new clustering in action yet, but I don't think constructing the facets would take too much time.  Certainly, the demo on the Vivisimo site is quick.
 
 
[1] http://www.exlibrisgroup.com/metalib.htm
 
[2] http://vivisimo.com/
 
 
--Dave

-------------------

David Walker
Library Web Services Manager
California State University
http://xerxes.calstate.edu

________________________________

From: web4lib-bounces at webjunction.org on behalf of Karen Coyle
Sent: Sat 12/30/2006 11:03 AM
To: kgs at bluehighways.com
Cc: web4lib at webjunction.org
Subject: Re: [Web4lib] Faceted navigation as metasearch



Karen, The problem with metasearch is that you're working with retrieved
sets -- and with the data in different formats. It would be great to
have some kind of standard set of facets and have those included as part
of the initial query reply (date, broad subject categories, etc.). I
know that the UC "Searchlight" metasearch was designed to pull back
retrieval numbers and display the numbers from each database that
replied. Think of how much more useful it would be to get more than just
a number. The records themselves wouldn't be retrieved until the user
initiates a display.

I think that pulling in all (or even some) of the retrieved records and
creating the facets on the fly would be prohibitively slow. Anyone know
a system that can do this?

kc

K.G. Schneider wrote:
> One idea I heard recently had to do with using faceted navigation search
> engines for metasearch. I've been mulling this over, particularly after
> rereading Karen Calhoun's "Changing Nature of the Catalog"
> (www.loc.gov/catdir/calhoun-report-final.pdf  ) and really, I'm trying to
> figure out if there's much wrong with this idea.
>
> Consider using a tool (Endeca, Siderean, FAST, i411, Dieselpoint-full
> disclosure: we implemented Siderean at my FPOW) that lies above the catalog
> and all other discovery services and provides access to ETDs, book data,
> journal articles, and more, ecumenically searched but parsed out logically
> in post-coordination. In fact, I'm wondering if NCSU and other libraries now
> on the faceted-navigation bandwagon are looking at moving in this direction.
>
> I do see one biggy. These search engines are themselves pretty
> lickety-split, but I wonder how slow retrieval would get if they were
> accessing separate journal and database services (versus what they do now,
> which is search their own pre-built indexes...). In some cases, it might be
> easy to create an index (e.g. if we have ETDs, I don't see why they can't be
> indexed). But journal articles? Pondering again... though if they were
> accessing resources such as Google Scholar and then tying everything back
> in...
>
> Karen G. Schneider
> Acting Associate Director of Libraries for Technology & Research
> Florida State University
> Email/AIM: kschneider at mailer.fsu.edu
> Blog: http://quodvide.wordpress.com
> Phone: 850-644-5214
> Cell: 850-590-3370
>
> _______________________________________________
> Web4lib mailing list
> Web4lib at webjunction.org
> http://lists.webjunction.org/web4lib/
>
>
>  

--
-----------------------------------
Karen Coyle / Digital Library Consultant
kcoyle at kcoyle.net http://www.kcoyle.net
ph.: 510-540-7596
fx.: 510-848-3913
mo.: 510-435-8234
------------------------------------

_______________________________________________
Web4lib mailing list
Web4lib at webjunction.org
http://lists.webjunction.org/web4lib/