DSpam filter

[The_Librarian]

Hi

Anybody else using the Dspam filter to filter out spam, pr0n, phishing mails and viruses?

Works great for me.

Regards

TU

[The_Librarian]

Update :

Sorry about bumping this thread to the top, but for those who have no idea what I'm waffling on about, here's a bit of information taken from the dspam (http://dspam.nuclearelephant.com/) website to whet your appetite.

Also, on the Smoothwall forums, you can find the thread here (http://community.smoothwall.org/forum/viewtopic.php?t=17691).

You might have noticed that they keep on changing the headers, sender and the such for the blue pills spam, and porn mails, this filter picks them all up... and legitimate emails is not blocked, unless by mistake on the part of the operator.

Even the Rolex watch advertisements is blocked.

Viruses (it works in conjunction with ClamAV) and phishing mails is also blocked.



DSPAM is an adaptive filter which means it is capable of learning and adapting to each user's email. Instead of working off of a list of "rules" to identify spam, DSPAM's probabilistic engine examines the content of each message and learns what type of content the user deems as spam (or nonspam). This approach to machine-learning provides much higher levels of accuracy than commercial "hodge-podge" solutions, and with minimal resources. DSPAM's best recorded levels of accuracy have included 99.991% by one avid user (2 errors in 22,786) and 99.987% by the author (1 error in 7000), which is ten times more accurate than a human being!


It also is very easy to use, no fancy rules or filters to set up. You simply mark the incoming spam as such, and it then "learns" which type of mail is spam and which is not. Because it does not use "static" filters, but statistical filtering, it can pick up spam even though the sender and header keeps on changing.

Spammers change their tactics in order to throw off most of the commercial mail filters, and this also take up a lot of your time...

Hope this helps you in the war against spam.

Regards

TU

[Oliver]

Thanks for this TU, it looks quite useful, I'll give it a try.

Cheers!