[Spambayes] Better default configuration for SpamBayes?
kenny.pitt at gmail.com
Mon Oct 31 19:37:52 CET 2005
On 10/29/05, Tony Meyer <tameyer at ihug.co.nz> wrote:
> >> The problem I have with this configuration is, that overtraining
> >> and wrong
> >> classification happen too easily. [...] perhaps you should change
> >> the
> >> default configuration of SpamBayes accordingly?
> > +1
> +1 from me, too. I've wondered about doing this in the past, but
> haven't got around to it (it is simple enough for users to change the
> default themselves, but it would be nice if they didn't have to).
Another +1 here. For anyone whose received mail is not balanced
between ham and spam, train-on-everything will quickly drive their
training to an imbalanced state as well, so it should probably be
discouraged for any inexperienced users (i.e. those who generally
stick to defaults).
More information about the SpamBayes