Print

Print


(please excuse cross-posting)

All,

If what follows seems *too enthusiastic*, forgive me.  One of my non-lib
friends who has published papers and recently told me how frustrated he
got with library catalogs, said in response to this tool: "Simply
amazing. Really" - and put it on his wiki page immediately, so I'm going
on that.  Also, James Weinheimer, a frequent poster on AUTOCAT and
NGC4Lib lists, has said "This is a wonderful beginning! It looks as if
there is finally something browsable for ordinary people to use... MUCH
easier for the public to search than the LC authority file" (of course
James wanted to rewrite this, fearing it sounded "a lot like a blurb on
a book cover", but I wouldn't let him, not wanting to lose the impact of
his initial enthusiastic reaction).

Before I myself go into "active promotion mode" please note: As Bernhard
says, *"This is still a demonstration of a concept rather than a
full-fledged and robust application.  After all, it was only a few hours
I could devote to it."*  He also adds that *"Improvements are certainly
necessary, but also possible",* adding that, "There are... imperfections
in the data conversion, as I left out all the 670s and such stuff that
is not of immediate use for searching. Also, some special characters are
wrong. And there are duplicates."  There are currently 5.3 million
records and 14.9 million index entries.  Scope notes will be
forthcoming.  

With that said...

Many of us have no doubt noted that Google Book Search has started to
use one part of librarians' inheritance -LCSH - no doubt to bring some
order to the chaos... to control the flood of information. 
(http://www.catalogingfutures.com/catalogingfutures/2007/10/google-using
-lc.html)

This makes sense.  Search engines like Google and the like, with their
undeniably useful mathematical algorithms, popularity ("relevance")
rankings, and ability to pinpoint exact phrases, etc. obviously have
their place - especially to obtain quick facts.  But we all experience
not only the power, but also the limitations of keyword searching.
Issues of the quality of results aside, search engines themselves are
*not always even the best way of getting quality information in a timely
fashion*, depending on one's area of inquiry (disciplinary focus),
purpose, etc.  And though we might make some serendipitous finds using
Google, etc., sometimes the amount of hits unrelated to our spoken and
unspoken questions can be very frustrating - what we really want is a
more "controlled" - and even "assured" type of serendipity *** - see
below  

Many of us also probably noted that back in Feb., LCSH authority data
was "freed" (and probably a lot of people were trying to figure out what
they could do with it)
(http://www.librarian.net/stax/1970/loc-authority-files-yours-to-keep/) 

Bernhard Eversberg (Universitetsbibliothek Braunschweig
[log in to unmask]) has done something - and I think his accomplishment
will sink in more and more, as you play with the tool: 

http://www.biblio.tu-bs.de/db/lcsh/ (screen with explanation)

http://www.biblio.tu-bs.de/db/lcsh/detail.php (boolean search option
screen)

Try it out.  I suggest the topic of World War II.  Start your search by
choosing "phrase" and using "ww II", "world war 2", or "world war two" -
and see what happens (and keep in mind, as James Weinheimer likes to
point out to his students, a keyword search for "world war II" is not
going to get you any primary documents :-) ).  

Also, on the Boolean search option screen try "battle" in the first
field and "world war" in the second, and press the "search" button...

Also... choose the "word" search and try "horses".  Browse down the list
and you will notice "Equine herpesvirus diseases", probably asking
yourself, "How did my search term 'horses' find that... how did it know
this was related?"  Click on it to get an idea of how it happened.  No,
it is not "semantic web" technology.  Rather, the whole valuable
structure of LCSH is simply being more effectively utilized and
presented in a more intuitive way for the user (and the links right into
WorldCAT, Google Books, and LT are pretty cool too).

As Bernhard puts it: "LCSH, in particular, consists mostly of multi-part
terms which to input correctly as such is outside reality. But the
complexity of it represents a considerable investment that can be tapped
into by browsing, with very little effort."

(Aside: I think this is a great complement to the Endeca-type of search
we've become accustomed to seeing, which takes a basic keyword search
and utilizes the structure of LCSH and LCC, dropping the user into a
structured environment that is easy to use, and can entice them to surf
around the catalog some more with the help of the controlled vocabulary,
structure, etc).   

Of course, useful "maps" of the world's persons, places, things, ideas,
etc. might come to us in a variety of forms.  Alphabetical lists of
things have been found to be particularly useful, if not necessary, for
collections of large size, and have been with us since the library of
Alexandria.  This, of course, does not preclude still other creative
ways of doing lists and browsing, perhaps by combining this text with
this related text or image, etc.  

Kevin Kelly of Wired magazine has predicted that users of the future
electronic library would be able to bring together, "all texts- past and
present, multilingual - on a particular subject".  I submit that
Bernhard has provided *a* *realistic* - and exciting - way forward for
us - showing how the old and the new can come together to serve the
user.  I suspect there are some real possibilities here.    

I think that all in all, this is not too bad for something done in one's
spare time without funding.  Play with the tool and let Bernhard know
what you think. 



***for more thought on the limitations of this kind of searching and
inquiry, see the fine articles referenced by these blog posts:  

-http://www.everythingismiscellaneous.com/2007/06/25/why-we-need-librari
ans/

-http://radar.oreilly.com/archives/2007/09/the_work_of_lea.html

-http://www.futureofthebook.org/blog/archives/2007/11/digitization_and_i
ts_discontents.html)

[sorry if all my URLS in this message break up - I can't get them to
work with tinyurl for some reason]

Regards, 

Nathan Rinne
Media Cataloging Technician
ISD 279 - Educational Service Center (ESC) 11200 93rd Ave. North Maple
Grove, MN. 55369 Work phone: 763-391-7183