On 8/28/06, Matt Mahoney <[EMAIL PROTECTED]> wrote:
Btw. does any current AI system which uses some probabilistic language model
infer that if two words can occur in identical contexts and affect
the distribution of later words pretty much identically they are pretty likely synonyms?
I would be interested in reading some related stuff so if any of you can point me to some non-trivial research on what can be inferred (and how) from such language models let me know.
However, I think that a lossless model can reasonably derive this information by observing that p(x, x') is approximately equal to p(x) or p(x'). In other words, knowing both x and x' does not tell you any more than x or x' alone, or CDM(x, x') ~ 0.5. I think this is a reasonable way to model lossy behavior in humans.
Btw. does any current AI system which uses some probabilistic language model
infer that if two words can occur in identical contexts and affect
the distribution of later words pretty much identically they are pretty likely synonyms?
I would be interested in reading some related stuff so if any of you can point me to some non-trivial research on what can be inferred (and how) from such language models let me know.
To unsubscribe, change your address, or temporarily deactivate your subscription, please go to http://v2.listbox.com/member/[EMAIL PROTECTED]
