RE: iHateSpam vs GFI

  • From: "Bill" <portabillemail@xxxxxxxxxx>
  • To: exchangelist@xxxxxxxxxxxxx
  • Date: Mon, 27 Oct 2003 14:53:51 -0700

We also are evaluating GFI ME9.  
I like the idea of the SPAM and HAM users, but my concern is with the
forwarding of the emails.  The act of forwarding the email causes the
original to be opened.  So, wouldn't that in effect cause any malicious
code in the email to be executed and making the situation worse?  I
actually have a SPAM user and I can copy messages from my mailbox to his. 
Then I could do the same operation you mention.
Has this been a problem?

> This is a multi-part message in MIME format.
> ------_=_NextPart_001_01C39A72.E652CF70
> Content-Type: text/plain;
>       charset="us-ascii"
> Content-Transfer-Encoding: quoted-printable
> Hi
> =20
> I'm evaluating GFI Mail Essentials right now. The latest version has
> Bayesian filtering which is a hughe step forward. At the moment Bayesian
> filtering is considerd one of the best methods to filter out spam BUT it
> does take some effort. Actually, the more spam you get, the better the
> filter gets.
> =20
> What I have done for my organisation:
> - Created a user SPAM . Told GFI to forward all spam to that user.
> - Created a folder called HAM and a folder called FALSE SPAM.
> =20
> I told my users to regularly check the SPAM folder and if they find
> e-mails that are not considered spam to them, they could drag it to the
> HAM folder. If in their normal mailbox they receive an e-mail which is
> spam but not considered spam by the filter, then they should move the
> mail from their mailbox to the HAM folder.
> =20
> Once every few days, I use the Bayesian wizard to update his settings
> according to the HAM and FALSE SPAM folders. And I must admit that we
> get less false positives every time.
> =20
> Since 14th oct, we have received only 2 false positives, 6 mails that
> were not marked as spam but should have been and over 500 SPAM mails.
> In the beginning it was much worse. So a Bayesian filter gets better
> every time you teach it.
> =20
> And now you should just look for software that has a great interface to
> your bayesian filter. Ofcourse, other filters are also needed, don't
> forget about them.
> =20
> Gabrie
> =20
> =20
> =20
> ________________________________


Other related posts: