Natural lang/expert systems/www searching

[NICK TOMAIUOLO, INSTRUCTION LIBRARIAN, CCSU]

Dear readers,
Not long ago P. Stephen Thomas made 
an astute observation was made by on the WEB4LIB list.  The 
comment was something to the effect that "Natural language" searching was
the "Holy Grail" of WWW search engines/services.  Among Mr. Thomas's 
cogent remarks was an allusion to artifical intelligence, which he
called Expert Systems, wherein searchers could pose a query and be 
intelligently prompted by the search services for further qualifying
information.

for example, Searcher: "Give me Supreme Court decisions" Engine: "From the
US or Canada?"  Or Searcher: What are the advantages of a flat tax?
Expert System: Do you wish to limit your topic to the United States?
Searcher: Yes.
Expert System: Would you accept disadvantages as well as advantages?
Searcher: Yes
And so on.

My question which may engender comment or discussion from not only 
professional searchers but also search engine developers, end-users, 
reference librarians from all fields as well as computer professionals is:

Does any such "Artificial Intelligence" powered engine exist anywhere for any
purpose?  In Library Reference, in Defense Industry, in Commerical business?
Again--I don't mean an engine that says "Narrow search by looking for
word in the title?" or the rudimentary type of search strategy, but one that
responds to the particluar query of the user.

Of particular significance would be **WWW** search engines that work 
intelligently-- that is, interact with the user.  Again, Are there any?  
Have you used them?  Where are they?

But, moreoever, does any AI engine exist in any realm that
intelligently prompts the user to facilitate focusing a query and therefore
making relevant retrieval more possible?  For example, even 
an interoffice database, a commercial database of some sort...

Regarding the topic of Natural Language searching, does anyone know of systems
or services or engines that can "translate" an end-user's query of 
for example
"find information on gender bias in compensation in professional sports" into 
its possible incarnations (i.e., "why are women paid less than men for winning
a golf tournament?")?

BTW -- would any medical information professionals, including readers at the
NLM be able to respond as to the status of the "Metathesaurus" (a project that
purported to take a query in the natural language of any professional and 
translate it into a coherent query for searching the NLM databases); does the
Metathesaurus do the intelligent translating for (at least) health information
related queries?  What system(s) are running with the Metathesaurus (e.g.,
NLM databases?  Anything beyond that?)?

Thank you for your comments,
Please note, I do not subscribe to several of the lists to which this is 
being posted.  If responding, would you send me a CC or simply respond to me
if appropriate?

Nicholas Tomaiuolo, MLS
Instruction Librarian
Central CT State University
New Britain, CT  USA
860-832-2068

Nicholas G. Tomaiuolo, MLS
Bibliographic Instruction Librarian
Central Connecticut State University Library
Reference Department
New Britain, CT  06050

email=  tomaiuolon at ccsu.ctstateu.edu
phone=  (860) 832-2068
fax=    (860) 832-3409