You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@spamassassin.apache.org by J4K <ju...@klunky.co.uk> on 2011/08/22 15:46:14 UTC
[Q] Bayes dB: ratio of spam and ham heavily in favour of ham
Afternoon gentlemen,
Seems the Bayes dB has become lop-sided in favour of ham. SA is
doing its job as there is little spam coming through these recently. I
had hoped we could keep it one third spam and two thirds spam. Does the
slant shown below (nspam verses nham) cause any problems with correct
Bayes classification of newly rated messages?
# sa-learn --dump magic
0.000 0 3 0 non-token data: bayes db version
0.000 0 640 0 non-token data: nspam
0.000 0 7001 0 non-token data: nham
0.000 0 366899 0 non-token data: ntokens
0.000 0 1085514555 0 non-token data: oldest atime
0.000 0 1314019965 0 non-token data: newest atime
0.000 0 0 0 non-token data: last journal
sync atime
0.000 0 0 0 non-token data: last expiry atime
0.000 0 0 0 non-token data: last expire
atime delta
0.000 0 0 0 non-token data: last expire
reduction count
Cheers, S
Re: [Q] Bayes dB: ratio of spam and ham heavily in favour of ham
Posted by Benny Pedersen <me...@junc.org>.
On Mon, 22 Aug 2011 15:46:14 +0200, J4K wrote:
> # sa-learn --dump magic
> 0.000 0 3 0 non-token data: bayes db
> version
> 0.000 0 640 0 non-token data: nspam
> 0.000 0 7001 0 non-token data: nham
> 0.000 0 366899 0 non-token data: ntokens
its not really possible to say if one nham is one ntokens, and same
goes for nspam, so in all its not possible to know if it bad learned
either, but keep monitoring it so says if rbl testing in mta is doing a
rate of bayes learn, test one spam, test one ham, if bayes agre then its
learned correct
basicly one should learn all msgs, not just the ones that is learned
bad, this keeps more tokens vaigthed good so false is reduced that way
bayes lives in a very dynamic world :=)