Newsgroups: comp.ai.doc-analysis.misc,comp.ai.nat-lang
Path: cantaloupe.srv.cs.cmu.edu!rochester!cornellcs!newsstand.cit.cornell.edu!ub!dsinc!netnews.upenn.edu!msunews!agate!howland.reston.ans.net!newsfeed.internetmci.com!in1.uu.net!bcstec!nntp
From: Richard Wojcik <rwojcik@redwood.rt.cs.boeing.com>
Subject: Re: Software Knows to Deconstruct in Plain English
X-Nntp-Posting-Host: snake.rt.cs.boeing.com
Content-Type: text/plain; charset=us-ascii
Message-ID: <312B54B2.28DA@redwood.rt.cs.boeing.com>
Sender: nntp@bcstec.ca.boeing.com (NNTP News Access)
Content-Transfer-Encoding: 7bit
Organization: Research & Technology
References: <4ftflg$st9@la1.digilink.net> <4fvchh$bcl@thor.cs.cornell.edu>
		<4g3656$gpt@la1.digilink.net>
		<TED.96Feb18222415@hellespont.crl.nmsu.edu>
		<4gdgnn$d2h@la1.digilink.net> <DOWDING.96Feb20182843@Gansett.ai.sri.com>
Mime-Version: 1.0
Date: Wed, 21 Feb 1996 17:21:54 GMT
X-Mailer: Mozilla 2.0b5 (X11; I; SunOS 5.4 sun4c)
Lines: 39
Xref: glinda.oz.cs.cmu.edu comp.ai.doc-analysis.misc:132 comp.ai.nat-lang:4565

John Dowding wrote:
> 
> >>>>> "Intelligent" == Intelligent Text Processing <reynolds@itpinc.com> writes:
> 
>     Intelligent> We have two types of people trying out the site.
> 
>     Intelligent> 1) Those who ask one run-on sentence that does not
>     Intelligent> fit in the query box and then post to the net saying
>     Intelligent> it doesn't work after it correctly answers their
>     Intelligent> question.
> 
>     Intelligent> 2) Those who ask 20 questions, get good results, and
>     Intelligent> call to see about incorporating the technology on
>     Intelligent> their site.
> 
> I see.  So, people who ask questions you can't answer are really just
> being uncooperative.

I've been following this thread with some interest.  The problem with AI
is that it can never live up to people's expectations, even when it does
some things spectacularly well.  I do not know how well Inquizit works in
real work environments, but I do believe that it will always handle some
queries very poorly.  The problem with testing an NLI is that you really
have to understand what it is for and who would reasonably want to use it.
Is it reasonable to ask a database questions that can only be answered by
the Oracle at Delphi?  Does the program fail because it can't make sense
of every question?  Perhaps Inquizit needs to fail more gracefully than it
does when it runs into a brick wall of a sentence.  The true test, however,
is whether it responds to the needs of users who understand the database
and what information they need to get out of it. Those users are likely to
make fewer goofball queries and also likely to be more forgiving of
occasional goofball responses.

-- 
Richard H. Wojcik, Natural Language Processing
Boeing Information & Support Services
P.O. Box 3707, MS 7L-43, Seattle, WA 98124-2207 (phone: 206-865-3844)
rwojcik@redwood.rt.cs.boeing.com                (fax:   206-865-2965)
----Opinions expressed above are not those of The Boeing Company----
