If someone puts a gun to your head and asks you to pick the most promising
approach, what would you say?
To me, it seems that an ideal system is a rule-based one that can learn
statistically.
I would favor "statistical rule induction", and indeed it might make
sense to seed the rule inducer with some hand-crafted rules, to guide
its learning in the right direction.
The innovation in my approach is to open the rules-base to the public.
Some people out there have a much better understanding of the complexities
of NL, and they don't need to be computational linguists to be able to
enter some rules, if the system is easy to use, though some training is
needed.
I really doubt that a collaboration of web-surfing NL enthusiasts are
gonna create better rules than the linguistics community has done so
far.
Why not just create a Web UI allowing users to enter additional rules
for some existing grammar, such as the Link Grammar (my personal fave)
or XTag (too complex for my taste)? I really think few people will
gain the needed skill and understanding to contribute. But we did add
some rules to the Link Grammar for a paid NLP consulting project a few
years back.
By the way, for the sub-goal of Basic English, a rule-based system may be
very adequate. And that's useful as a first NL interface.
Basic English is not all that unambiguous. Sentences may be short but
anaphora and prepositions remain.
If you're going to restrict your AI to a special subset of English,
then it can't read free text anyway... all you can do is chat with it.
So why not just chat to it in Lojban which has full expressive power,
barely ambiguous semantics, and totally unambiguous syntax?
-- Ben
-----
This list is sponsored by AGIRI: http://www.agiri.org/email
To unsubscribe or change your options, please go to:
http://v2.listbox.com/member/?member_id=231415&user_secret=fabd7936
