Hello Loren,

Saturday, February 28, 2004, 1:49:48 AM, you wrote:

LW> Going through a few spams tonight, I came up with a small handful of
LW> possibly useful rules.
LW> I'd appreciate it if anyone who has a corpus set up and some spare cpu
LW> cycles could test these and see if they are actually any good.

LW> Thanks,
LW>         Loren

Results:

(First numeric frequencies, followed by percentage frequencies)

OVERALL     SPAM      HAM     S/O          SCORE  NAME
 106532    87317    19215    0.820   0.00   0.00  (all messages)
   7595     7595        0    1.000   1.00  10.00  PT_WORDLIST_30
    244      244        0    1.000   0.99   1.00  DES_ENC_STMP
   1665     1661        4    0.989   0.95   1.00  X_BOGUS_MAILER
   1653     1647        6    0.984   0.93   0.50  BOGUS_SUBJECT
    345      342        3    0.962   0.86   1.00  X_UNAUTHENTIC_WARNING
      0        0        0    0.500   0.00   0.10  BOGUS_MSGID
      0        0        0    0.500   0.00   1.00  STMP_NO_ID

OVERALL%   SPAM%     HAM%     S/O    RANK   SCORE  NAME
 106532    87317    19215    0.820   0.00    0.00  (all messages)
100.000  81.9632  18.0368    0.820   0.00    0.00  (all messages as %)
  7.129   8.6982   0.0000    1.000   1.00   10.00  PT_WORDLIST_30
  0.229   0.2794   0.0000    1.000   0.99    1.00  DES_ENC_STMP
  1.563   1.9023   0.0208    0.989   0.95    1.00  X_BOGUS_MAILER
  1.552   1.8862   0.0312    0.984   0.93    0.50  BOGUS_SUBJECT
  0.324   0.3917   0.0156    0.962   0.86    1.00  X_UNAUTHENTIC_WARNING
  0.000   0.0000   0.0000    0.500   0.00    0.10  BOGUS_MSGID
  0.000   0.0000   0.0000    0.500   0.00    1.00  STMP_NO_ID


Bob Menschel


Reply via email to