[Corpora-List] Quotable Statistics on Unstructured Data on the WWW

[John F Sowa]

On 6 Dec 2013, at 17:10, Stefan Bordag <bordag at exb.de> wrote:
> Even using the best full text capable search engine you are never guaranteed
> to get the one information you were looking for, and only that information.

On 12/6/2013 12:51 PM, Trevor Jenkins wrote:
> As someone who worked in R&D and post-sales support of the world's best
> free text search engines (3rip and Trip) for nearly three decades
> I dispute that claim.

I agree with everybody's complaints about every existing search
and query system -- whatever it may be called.

But I also believe that even intelligent humans who are familiar with
the subject matter miss or misinterpret a huge amount of relevant and
even critical material.

As evidence, just look at the failures "to connect the dots" by agencies
that have the best computing resources and highly trained personnel.
After a disaster, "post mortems" (literally) can get better answers
-- primarily because they have a better idea of what to look for.

I would not call human retrieval "the gold standard".  Trained humans
with good tools might reach a "copper standard".  But I believe that
much more R & D is needed to develop tools that can enable trained
humans to reach a "silver standard".

John

_______________________________________________
UNSUBSCRIBE from this page: http://mailman.uib.no/options/corpora
Corpora mailing list
Corpora at uib.no
http://mailman.uib.no/listinfo/corpora