RE: Corpora: Ergo's Patent Publishes

Philip A. Bralich, Ph.D. (annes@htdc.org)
Tue, 16 Mar 1999 17:24:40 -1000

At 09:46 AM 3/16/99 +0000, you wrote:
>Congratulations. Your refutation is as simple and eloquent as it is damning.
>Having spent time in the past working in the area of parsers and theories of
>syntax, I usually come to these claims with the thought: who needs syntax?
>Put less bluntly, if a theory of grammar is to have any value at all in a
>piece of computer software, its use needs to be argued from a design point
>of view; it also has to pass evaluation tests which demonstrate its fitness
>for this design purpose. Paradoxically it seems that Bralich may be making
>the same point, if inadvertently. Though in the case of his parser, I don't
>see the purpose.
>

I recommend that you look at the sentences proposed in the parsing contest
on our web site. They are focused on demonstrating the value of a theory of
syntax in a particular domain -- most notable being current speech recognition
devices and the improvement in the size, scope, and amount of commands
that can be handled through the use of a theory of syntax. It is only by
looking at specific practical domains and then comparing parsers side by
side with the same sentences that the value can be seen. Take a few minutes
and have a look at the stuff under "Parsing Contest" on our web site at
http://www.ergo-ling.com. You will find that I am making the same point
and that I am making it with working software in practical real world (i.e.
potential jobs) rather than with a few miscellaneous sentences whose import is
questionable.

Phil Bralich

Philip A. Bralich, Ph.D.
Ergo Linguistic Technologies
2800 Woodlawn Drive, Suite #175
Honolulu, HI 96822
fax: (808)539-3921
tel: (808)539-3924