Printable View
I currently use Mailwasher or XTerminator as my Spam filters as they only download the headers and can delete the emails from the server.
This saves me time when I am on dialup and getting loads of spam.
Recently I have been reading more about Bayesian filters and how they are better than black/white lists.
Are there any free Bayesian filters that don't have to download the whole email?
Or is it necessary by their very nature to need to download the emails before processing them?
Not 100% sure that you will find a free one but they work great. We run GFI Mail Essentials for Exchange Server here at my work and
the Bayesian spam folder gets about 95% of the spam. The other spam folders get the remailing 5% (Blacklist, DirHarvest, DNSBlacklist, Header,
Keyword, SPF, & SURBL). I cant remember the last time I got spam in my Inbox folder.
:p Got Spam!
I use a Bayesian spam filter called PopFile. It is a free, open source program and it works very well. Once trained I rarely get false positives and it correctly identifies 95% of spam. It even seems to be effective at identifying the spam which contains word salad.
The nature of the Bayesian filter does however mean that the entire email needs to be downloaded and analysed and unless the email can be identified as spam from the headers alone other spam filters need to do the same.