On Wed, 2010-08-11 at 17:58 +0530, JAGANADH G wrote: > > after some years of experimentation, I find a setting of 3.5 on > > spamassassin along with some training eliminates 99% of spam - and > very > > rarely catches non spam. > > - > > > It is a statistically trained system to determine weather a mail > content is > spam or non spam. > I think it is having a bootstrat facility to learn new spam mails too. > It may be the reason of its slow performance.
this can be adjusted - after training is over, there are many ways of speeding it up and manually adjusting the rules. -- regards KG http://lawgon.livejournal.com _______________________________________________ ILUGC Mailing List: http://www.ae.iitm.ac.in/mailman/listinfo/ilugc
