On Mon, 2007-09-03 at 23:39 -0400, Michael Fitzgerald wrote:
> At 11:06 PM 9/3/2007, Conal Tuohy wrote:
> >In the "book space", I would expect to see e.g. Amazon and
> >Google deploy these techniques first, for disambiguating authors,
> >identifying concepts, etc, simply because they have access to the full
> >text of books, and we don't.
>
> But as I thought I demonstrated a short while ago (late May), Amazon
> at least has apparently no interest in disambiguating authors,
Well ... APPARENTLY not only since they haven't done it YET, but that
doesn't mean that
* they aren't working on it,
* they have no interest in it, or
* they won't ever DEVELOP an interest in it
> and
> library catalogers do - and they have done so (even without access to
> the full text - meaning that the job can be done right now if one
> cares enough to bother).
Of course it can be done by humans without access to electronic full
text, but the techniques I referred to are based on machine learning and
require electronic full text.
> Amazon isn't in the same business as
> libraries. This difference in priorities is quite likely to result in
> different handling for situations. If you spend more at the store do
> you really think the shopkeeper is going to worry about whether you
> got the most efficient or most complete view of things? Ditto
> regarding the eyeballs on Google's advertisements.
I don't see why Amazon's site wouldn't also benefit from better
authority control? Surely if it makes their site more usable then that
will be a net benefit to their business? NB they ARE already doing some
computationally-flash stuff with that full text - namely identifying
"Statistically Improbable Phrases":
http://www.amazon.com/gp/search-inside/sipshelp.html/002-7908410-9151243
e.g.
http://tinyurl.com/32lbjm
--
Conal Tuohy
New Zealand Electronic Text Centre
www.nzetc.org
Received on Tue Sep 04 2007 - 00:36:04 EDT