Dan Brian wrote:
>
> David L. Nicol wrote:
>
> > Dan Brian wrote:
> > >
> > > disambiguation statistics
> > >
> > > Dan
> >
> > Statistics break at the edges. I meant something that
> > will expand
>
> I know, just throwing in the ambiguity bit. The code I'm talking about
> expands:
>
> "Please open the site www.perl.com and print out the content."
>
> into the appropriate code to be evaled.
I am reminded of the research on NLP I looked at for philosophy of
language class. There was a chapter of noun-noun-noun sequences
in english.
There is nothing, syntactically, separating "bird feeder kit" from
"child library card" but one of them binds ((bf)k) and the other
binds (c(lc)) for instance.
Of course, you can try to drop the word-break distinction as being
arbitrary, which is fun to think about for a while, but creates
a high barrier to continued discussion
--
David Nicol 816.235.1187
"Perl was born in downtown Hell!" -- Matt Youell
