efa-project.org

Hi

I have noticed that all my emails are marked as BAYES_00 even when the email was clearly spam, see the scores below;

Is there something I should be doing so every email is not given a -1.90 score by default? I reguarly feed it ham and spam via"Reports >> Message Operations"

Thanks Will

-1.90 BAYES_00 Bayes spam probability is 0 to 1%
1.10 DCC_CHECK Detected as bulk mail by DCC (dcc-servers.net)
0.29 DIGEST_MULTIPLE Message hits more than one network digest check
0.00 HTML_FONT_LOW_CONTRAST HTML font color similar or identical to background
0.00 HTML_MESSAGE HTML included in message
5.00 KAM_VERY_BLACK_DBL Email that hits both URIBL Black and Spamhaus DBL
1.39 PYZOR_CHECK Listed in Pyzor (http://pyzor.sf.net/)
0.50 RAZOR2_CF_RANGE_51_100 Razor2 gives confidence level above 50%
1.89 RAZOR2_CF_RANGE_E8_51_100 Razor2 gives engine 8 confidence level above 50%
0.92 RAZOR2_CHECK Listed in Razor2 (http://razor.sf.net/)
3.33 RCVD_IN_SBL_CSS Received via a relay in Spamhaus SBL-CSS
0.50 RCVD_IN_SORBS_SPAM SORBS: sender is a spam source
0.79 RDNS_NONE Delivered to internal network by a host with no rDNS
-0.00 SPF_HELO_PASS SPF: HELO matches SPF record
0.01 T_REMOTE_IMAGE Message contains an external image
1.70 URIBL_BLACK Contains an URL listed in the URIBL blacklist
2.50 URIBL_DBL_SPAM Contains a spam URL listed in the DBL blocklist

Are you training your bayes?

I am via message operations, is that correct?

Yeah, but it takes *a lot* of sample messages to influence bayes scoring, and doesn't happen overnight.

I have been running it for over a year? I will keep on feeding it and see what happens

Will

Any joy?

One problem I find with find with the bayes filter is that the spammers are finding new ways to defeat it temporarily until the filter is retrained which is why I cannot just rely on it solely.