Search engine question

[Richard Wiggins]

Bernie,

Interesting question. How literally do you mean the term "search engine"?  Do you mean a robot that crawls and indexes Web sites?  Something that you then search, as opposed to browsing? Most engines are robotic and sweeping in their search.  I'm not aware of anyone "training the spider" a la Findlaw.com for library science.  But various engines can filter for a given topic area pretty effectively.  

If you go into the Google Directory and limit your search to the area of  

   Reference > Libraries > Library and Information Science

... and then do your search, you have sort of simulated a Findlaw.com for library science.  Google is leveraging the same Open Directory hierarchical site list that Matt refers to.   So in this case a search on "catalog" brings up a highly relevant hit list.  Doing a search for that common word at the root of Google brings in far more irrelevant hits.

In Northern Light, the Custom Search Folders feature disambiguates between relevant topical areas (Library cataloging & classification) and not so relevant ones (Unidentified flying objects (UFOs)).  Here you can search at the NL home page and begin disambiguating on the first results screen.

The new kid Teoma also breaks things into subcategories, this time using hubs and authorities, not a taxonomy and automated classification.  It seems to do pretty well disambiguating "catalog" but to me NL's breakdown is more effective.

But:

1) I don't think there's a single winner here.

2) None of this uniquely applies to library science.  You could use similar techniques for searches in road construction.

3) The real question is how much of a discipline's literature is available on the public, open Web.  NL's Special Collection lets you get beyond that.  Then you'd have to ask how much of the library science literature is in their collection, and how much you're willing to pay for it.

This wasn't a trick question, was it?  :-)

/rich


On Thu, 13 September 2001, Matt Theobald wrote:

> 
> see
> http://dmoz.org/Reference/Libraries/Library_and_Information_Science/
> 
> 85 resources searchable
> also other categories available with single digit entries
> 
> -Matt
> 
> At 9/13/01 01:44 PM, Sloan, Bernie wrote:
> 
> >I think I need to clarify:  I am interested in search engines that would be
> >good for locating library and information science resources about the
> >profession, not library resources per se.
> >

Richard Wiggins
Writing, Speaking, and Consulting on Internet Topics
rich at richardwiggins.com       www.richardwiggins.com