[Web4lib] AI answering, IBM Watson, whatever...

Thomas Bennett bennetttm at appstate.edu
Thu Feb 10 09:57:19 EST 2011


On Wednesday 09 February 2011 20:28:09 Ernest Perez wrote:
> Howdy, Webbers,
> 
> I'd made my original post on the tail end of the IBM Watson/AI-answering
> topic in hopes of hearing more reactions about, "Exactly what is it that we
> want a mythical Answer Wizard to do, anyway?"

^
Yes and your post would make sense had it been Google, Yahoo, Ask Jeves, et 
al.. That process one query and can return thousands of answers that might be 
appropriate for the query.  Watson has to return the one response that makes 
sense.  Watson is distinctly different in that it is the reverse engineering 
what those search engines perform.  Not only to assemble the correct 
information having been given an answer but to pose that response in the form 
of a question that makes sense for the given answer.

fwiw

Thomas

> 
> I offered the information about my company's Web service, and their
>  approach from the view of linguistic analyzing, summarizing, condensing,
>  and organizing pertinent information relevant to a user query; of giving
>  the user the ability to stop when enough and appropriate information had
>  been delivered.
> 
> I have heard from one nearby academic library staff, where I asked the
> Reference Director to examine our application. He responded that he and 5
>  of the senior Reference staff tried out "iResearch Reporter," and felt
>  they had to give me negative feedback.
> 
> Their comments:
> 
>    - Felt that this made it "too easy" for the user.
>    - That it would not be effective in teaching them evaluation of
>    resources, in the proper methods of research, etc.
>    - That it was questionable, maybe a bit like "buying a college degree."
>    - That it didn't seem like a proper approach.
>    - That because of this, they ethically couldn't give a positive reaction
>    to this kind of text-mining application.
> 
> Whoa! Am I missing something here? I first associated with development of
> our product after my long library career of helping people and
>  organizations find information, answer questions, solve problems, etc. Or
>  as Mike Keonig has phrased it, "Putting something away and being able to
>  find it again."
> 
> In the past, I'd published two separate articles in Online Magazine about
> this kind of software technology, describing this linguistic analysis
> approach, and outlining the benefits as I saw them.
> 
> Coming from that view, I was surprised by the reactions of these six
> academic librarians. I thought the purpose of what we've all been doing was
> helping people effectively get the information we need.
> 
> 
> Surprise...! Perhaps we may all want to raise ethical objections to such
> questionable smoke & mirrors as IBM STAIRS (& Watson), Dialog, BRS, Google,
> Bing, Information Access, Ebsco, etc.      8-)
> 
> Can't make things too darned easy, after all. Eh?
> 
> Cheers,
>   --ernest
>  ---------------------------
> Ernest Perez, Ph.D.
> Power Text Solutions, Inc.
> http://www.irr-usa.com
> ernest at irr-usa.com
> _______________________________________________
> Web4lib mailing list
> Web4lib at webjunction.org
> http://lists.webjunction.org/web4lib/
> 

-- 
==========================================
Thomas McMillan Grant Bennett           Appalachian State University
Operations & Systems Analyst            P O Box 32026
University Library                                Boone, North Carolina 28608
(828) 262 6587

Library Systems Help Desk: https://www.library.appstate.edu/help/
==========================================




More information about the Web4lib mailing list