Ahh.. OK, thanks.  I could have sworn .6 was optimal.  Good to know.  

Paul K. Dickson
Systems Administrator III
Frederick County Government, I.I.T.
301-600-2399


-----Original Message-----
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Kevin
Sent: Thursday, March 27, 2008 6:01 PM
To: ASSP development mailing list
Subject: Re: [Assp-test] rebuild 'norm' odd jump.

Dickson, Paul wrote:
>  My understanding is the closer to .6 it is, the more accurate the
> bayesian filter will be. 

That is incorrect.
The norm should be above 0.6 and below 1.4 to be considered "healthy".
The ideal norm is 1.0000, which is an even mix of spam and not-spam
weights.

> 
> Found 12836215 spam words, 12238641 non-spam words.
> 
> norm=1.0488

This is a perfectly healthy corpus.

If you are getting false positives perhaps you should remove any 
spam/not-spam reports that are over a year old.

Kevin


------------------------------------------------------------------------
-
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services for
just about anything Open Source.
http://ad.doubleclick.net/clk;164216239;13503038;w?http://sf.net/marketp
lace
_______________________________________________
Assp-test mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/assp-test

-------------------------------------------------------------------------
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services for
just about anything Open Source.
http://ad.doubleclick.net/clk;164216239;13503038;w?http://sf.net/marketplace
_______________________________________________
Assp-test mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/assp-test

Reply via email to