2005-03-11 17:00:01 +01:00
|
|
|
Bayesian Noise Reduction is a statistical approach to evaluating coherence by
|
|
|
|
|
instantiating a series of machine-generated contexts to serve as a means of
|
|
|
|
|
contrast. This makes it possible to identify text that is out of context using
|
|
|
|
|
a form of pattern consistency checking. BNR attempts to solve the problem
|
|
|
|
|
commonly referred to as "Bayesian Noise" which, in its simplest definition,
|
|
|
|
|
refers to irrelevant data present in a message being classified. Bayesian Noise
|
|
|
|
|
Reduction dubs irrelevant text in order to provide cleaner classification and
|
|
|
|
|
is implemented as a pre-filter to existing language classification functions.
|
|
|
|
|
|
2008-06-01 11:42:37 +02:00
|
|
|
BNR is used in Dspam (mail/dspam, mail/dspam-devel - the ports don't depend on
|
2005-03-11 17:00:01 +01:00
|
|
|
this one)
|
|
|
|
|
|
|
|
|
|
See www for white-paper and presentation.
|
|
|
|
|
|
|
|
|
|
WWW: http://www.nuclearelephant.com/papers/bnr.html
|
