[clug] Spambayes

Felix Karpfen felixk at webone.com.au
Fri May 23 08:18:58 EST 2003

Nemo -earth native- wrote:
> I'm increasingly finding I'm putting thought into this... SO, straw poll
> time. 
> Q: What spamfiltering system do people use?
> Q: Is it trainable?
> Q: If it is, do you update it with ALL messages once classified as
> spam/ham, or only with false pos/neg results? (or not at all? Or
> something else?)

Two cents worth from an amateur.

I have been using a Canadian program (called "spamfilter" - no longer
maintained) for the last two years and it does a reasonably good job of
identifying spam.  Recently, I have added bogofilter so that I now use
"spamfilter" to look only at the email headers and "bogofilter" to look
at bodies of messages with inoffensive headers.

Looking at the track-records, it turns out that more than 80% of my
spam-mail has a Content-Type of either "text/html" or

I am now using "mailfilter" to delete from my ISP's mailserver all
emails not sent by any of my regular correspondents and having the above
Content-Types. And, as a result, the loads on my spamfilter programs has
dropped significantly.  The output of today's log of mailfilter-deleted
messages is attached, for information.   

I also update bogofilter daily with the following routines:

  bogofilter -s < /home/felixk/mail2001/IN-zzzzz 
  rm  /home/felixk/mail2001/IN-zzzzz 

"IN-zzzzz" is the box into "spamfilter" sends mail that it regards as

I hope that this helps the straw poll.  If needed, I could supply the URLs
for the programs that I use; but Google would find the URLs faster than
I can by looking on my box.

Felix Karpfen 

Felix Karpfen
felixk at webone.com.au
Public Key 72FDF9DF (DH/DSA)

-------------- next part --------------
mailfilter: 0.4.0 querying pop.webone.com.au on Fri May 23 06:59:14 2003
mailfilter: Examining 26 message(s).
mailfilter: Deleted Nancy Heidepriem <Celiaiwq at email.cz>: Wow em this summer...start now felixk, Wed, 21 May 2003 13:16:14 -0700. [Applied filter: '^Content-Type: text/html']
mailfilter: Deleted Trudy Waldron <Chereevp at mailserver.dk>: Equal results just cheaper, Wed, 21 May 2003 15:20:07 -0700. [Applied filter: '^Content-Type: text/html']
mailfilter: Deleted Lauretta Yount <Carlenasbm at freemail.globalsite.com.br>: Virile, Wed, 21 May 2003 17:21:50 -0700. [Applied filter: '^Content-Type: text/html']
mailfilter: Deleted Jeane Fralick <Luanneltf at 126.com>: , Wed, 21 May 2003 18:17:48 -0700. [Applied filter: '^Content-Type: text/html']
mailfilter: Deleted "Angelina Tapia" <9a9iimlktkm at msn.com>: Fwd: Use your computer to make money, Thu, 22 May 03 07:45:20 GMT. [Applied filter: '^Content-Type: multipart/alternative;']
mailfilter: Deleted Vincent Galvan <jason_fix at vampiress.zzn.com>: Balt, Thu, 22 May 2003 01:43:33 -0500. [Applied filter: '^Content-Type: text/html']
mailfilter: Deleted "" <MasterAdvertiser3193083 at hotmail.com>: Rates are predicted to go back up, REFINANCE Today! m, Fri, 23 May 03 02:19:41 GMT. [Applied filter: '^Content-Type: multipart/alternative;']
mailfilter: Deleted "Florence Conn" <siqs25yy2xr at usa.net>: wicosecond dulky bzegokocsqacd xtrg, Thu, 22 May 03 14:35:46 GMT. [Applied filter: '^Content-Type: multipart/alternative;']
mailfilter: Deleted "Yong Kaiser" <yong_kaiser at chello.at>: contact me back asap   uxsp1ujtnn, Thu, 22 May 2003 18:54:27 +0000. [Applied filter: '^Content-Type: text/html']
mailfilter: Deleted Margot <Sachikolxw at shoppersville.net>: Enjoy life more, Thu, 22 May 2003 08:28:12 -0400. [Applied filter: '^Content-Type: text/html']
mailfilter: Deleted "Blake Dalton" <220d5i00v at aol.com>: )Be Refinance free... Free online quote  z, Thu, 22 May 03 12:02:28 GMT. [Applied filter: '^Content-Type: multipart/alternative;']
mailfilter: Deleted "Lorrie" <GenericViagra319303 at hotmail.com>: fwd: have you seen this site?                   (ID:QVMU), Thu, 22 May 2003 22:58:36 +0800. [Applied filter: '^Content-Type: text/html']
mailfilter: Deleted "Ellis Leslie" <rwhdtb51 at aol.com>: For Home busines people xdwpwdw tzof, Thu, 22 May 03 13:01:56 GMT. [Applied filter: '^Content-Type: multipart/alternative;']
mailfilter: Deleted "Winifred Padgett" <cwl6wy644 at yahoo.com>: Soma prescribed online for free, Shipped Overnight, Fri, 23 May 03 05:27:11 GMT. [Applied filter: '^Content-Type: multipart/alternative;']
mailfilter: Deleted "Kay Samuel" <08t2bel4dc at yahoo.com>: Fw:Thanks For The Memory, Thu, 22 May 03 05:51:40 GMT. [Applied filter: '^Content-Type: multipart/alternative;']

More information about the linux mailing list