[Web4lib] RE: Federated search products and Full Text/Peer Reviewlimiting

Dale Askey daskey at ksu.edu
Fri Apr 21 11:58:59 EDT 2006


Can't just leave Julie's response unrebutted ...

Your comment is what I hear every time I bring up the deficiencies of 
fed searching within earshot of a fed searching vendor. Yes, in theory 
it is possible to do all sorts of fun things with inbound data on the 
client side. I nearly blinded myself writing elegant regular expressions 
to get a federated search product to morph the incoming data into shape 
to do all sorts of things.

My point is that there's a gap between the "anything is possible" theory 
and library reality. Any large research library has scores, if not 
hundreds, of databases that they offer which do not lend themselves to 
convenient federated searching. Vendors tend to ignore these (once 
you've purchased the product, that is), so you're on your own to write 
parsing programs and extra features to get them to work. It's hard and 
tedious work, and unless you've got a bunch of bored programmers--most 
libraries don't--then lights out on that. If you can't federate it, 
things like filters for peer-reviewed articles don't really mean a whole 
lot.

It's not a question of the possible, it's about the doable, and given 
the iconoclastic nature of some database vendors, well, it's more about 
lobbying than any technical issue.

Dale

Julie Blume Nye wrote:
>> one of the core weaknesses of 
>> federated searching, i.e.- it can't do a better job than (or even as 
>> good of a job as) any given database's native interface. Unless the 
> 
> Not necessarily true -- if the targets don't support a feature, the
> federated search software may be able to implement it on the client side.  
> 
> For example,  Fretwell Downing's ZPORTAL limits to peer reviewed journals by
> comparing results received from each database against a 'master' list of
> peer-reviewed journals.   (The comparison is on ISSN, not journal title.)
> This allows the peer reviewed limit to applied to results from many more
> databases since most databases include ISSNs in their results, while
> relatively few support peer-reviewed limiting at present. 
> Either Openly.Informatics' UHF data or Bowker's Ulrich's dataset can be used
> as the master list of peer-reviewed journals.  
> 
> This is also the approach ZPORTAL has taken for other features that many
> targets don't support (or don't support in comparable ways) -- e.g. date
> limiting.
> 
> ---------------------------------------
> Julie Blume Nye
> Senior Product Designer -- Fretwell-Downing, Inc.
> 407 River Trace Dr.
> Rougemont, NC 27572
> Phone: 336-364-2607 Fax: 336-364-4224
> E-mail: mailto:julie.nye at fdusa.com 
> Website: www.fdusa.com    
> 
> ~~~~~~~~~~~~~~ DISCLAIMER ~~~~~~~~~~~~~~ 
> The information transmitted in this electronic mail message may contain
> confidential and or privileged materials. For full details and restrictions
> see http://www.fdgroup.com/emaildisclaimer.html
> 
> 
> 
>> -----Original Message-----
>> From: Dale Askey [mailto:daskey at ksu.edu] 
>> Sent: Tuesday, April 18, 2006 12:53 PM
>> To: web4lib at webjunction.org
>> Subject: re: [Web4lib] Federated search products and Full 
>> Text/Peer Reviewlimiting
>>
>> Jonathan,
>>
>> I think your question gets at the heart of one of the core 
>> weaknesses of 
>> federated searching, i.e.- it can't do a better job than (or even as 
>> good of a job as) any given database's native interface. Unless the 
>> native interface allows limiting to full text or 
>> peer-reviewed articles, 
>> there's at best dim hope that you could get the former to work, and 
>> practically none for the latter. Now I know a few of you 
>> highly-skilled 
>> programmer types are just looking to prove me wrong, and for a few 
>> limited database targets you might succeed, but let's be 
>> realistic. With 
>> a remotely-hosted service, you're never going to have such 
>> control over 
>> the search behavior, and with locally-hosted, you'd need some 
>> very, very 
>> talented folk (read: large budget or blind luck) to get a fed search 
>> engine to do this. It may have less to do with the technical aspects, 
>> frankly, and more to do with the potpourri of data that the 
>> targets return.
>>
>> Besides, if you have a good link resolver, why limit to full text 
>> results? Just slap a link resolver button on each result, and that 
>> problem is largely solved.
>>
>> Dale
>>
>>
>>
>> quoted message:
>>
>> Our library is considering adding a federated search product, 
>> and we've
>> got a question for those of you now using them.
>>
>> In your experience, have Federated Search programs been able 
>> to reliably
>> limit to Full Text and/or Peer Reviewed articles?  We're especially
>> interested in those programs that are hosted remotely rather than
>> locally.
>>
>>
>> -- 
>> Dale Askey
>> Web Development Librarian
>> KSU Libraries
>> 118 Hale Library
>> Manhattan, KS 66506
>> (785) 532-7672
>>
>>
> 
> 
> _______________________________________________
> Web4lib mailing list
> Web4lib at webjunction.org
> http://lists.webjunction.org/web4lib/
> 

-- 
Dale Askey
Web Development Librarian
KSU Libraries
118 Hale Library
Manhattan, KS 66506
(785) 532-7672


More information about the Web4lib mailing list