[Mimedefang] SpamAssassin 2.40 experiences?

Clayton, Nik [IT] nik.clayton at citigroup.com
Thu Sep 5 07:43:01 EDT 2002


> Has anyone heard any more about the Baynesian analysis that 
> was all the rage a few weeks ago?

You can barely browse the web at the moment without stumbling across
some college grade programmer who's implemented a version of it.

That sort of approach is OK if you are either:

   a) Only filtering e-mail for one recipient, or

   b) Filtering mail for lots of recipients, but they all receive
      very similar e-mail

If you have lots of recipient who get different sorts of messages (say,
you're doing this at a bank, and a bunch of your users get financial
information by e-mail, and a bunch don't) a simple Baysian implementation
doesn't work.

Even if (a) and (b) do hold, the rules typically need to be trained on
a per-user basis with a spam/non-spam corpus as necessary.  That's 
probably feasible at small organisations, but it's not practical at 
larger ones.

N
-- 
1        1         2         3         4         5         6         7    7
         0         0         0         0         0         0         0    5
  -- The 75 column-ometer
                                             Contributing to the heat death
Global Messaging, 120 Cheapside, x83331      of the universe since 1973.



More information about the MIMEDefang mailing list