LOL!  I love your example.  :-)

I used to work at a company working on natural language processing (in
Smalltalk no less).  We had more than a dozen doctorate linguists and
computational linguists working at LingoMotors.  Here's just one single and
overwhelming example of a challenge to overcome.  A perfectly grammatical
sentence in a human language can have many valid parse trees, and realize
that this isn't a design fault of the parser.  Then you have to pick the one
that the speaker intended. This is no mean feat.

So, first correctly recognize all the spoken words (hard enough), being sure
to know where the sentence boundaries are (also hard), then parse them
correctly into the possible correct senses (much harder), and then finally
decide based on expert knowledge and context that may not be present which
sense is the correct one (really, really difficult).

Natural language wins?  Not anytime soon.

-Carl Gundel

-----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf Of
Casey Ransberger
Sent: Tuesday, April 09, 2013 4:49 AM
To: Fundamentals of New Computing
Subject: [fonc] When natural language fails!

Here's my example. 

Siri: Intruders detected on the tenth floor. 

Me: Okay Siri, seal off decks six through twelve. Open the airlocks. Number
one! Arrange a security detatchment, let's light a fire under their asses!

Siri: Aye captain. Retracting cooling rods from primary and secondary
reactors. Fire should commence within minutes. 

Me: No no no no! Put the cooling rods back into the reactors, Siri! What the
[explitive deleted] were you thinking??

Siri: Got it. I've made an appointment with your dentist for Monday.
Approximately three minutes fourteen seconds to meltdown in primary and
secondary reactors. Your dentist says hello, by the way. 

(etcetera)

The computer is going to keep getting smaller. How do you program a phone?
It would be nice to be able to just talk to it, but it needs to be able --
in a programming context -- to eliminate ambiguity by asking me questions
about what I meant. Or *something.*

It's tragic that Siri can't tell me what you get when you multiply six by
nine. I think it's been crippled, based on stuff Woz has said about what
Apple did when they bought it up. 

So there are some really interesting angles without well understood
solutions wrt NLP (and of course the group is welcomed to slap me in the
face with my ignorance because I know there's stuff I don't know.)

The best thing I've found to a natural language programming system is Inform
7 which leaves things to be desired. At least it is unambiguous, but I think
that in natural language, what we need are ways to cope with disambiguation.


Anyone want to point me at cool stuff to read? :D

-- Casey Ransberger


_______________________________________________
fonc mailing list
[email protected]
http://vpri.org/mailman/listinfo/fonc

_______________________________________________
fonc mailing list
[email protected]
http://vpri.org/mailman/listinfo/fonc

Reply via email to