http://bugzilla.spamassassin.org/show_bug.cgi?id=3175
------- Additional Comments From [EMAIL PROTECTED] 2004-03-15 09:35 -------
The following are working well for me. Shorter versions are trapping on far
too much ham.
Contrary to one of your assumptions, in the last few weeks poison blocks have
started to include punctuation at reasonably appropriate statistical intervals
to defeat the 'no punctuation' tests. They are still failing the other checks
though, so are still reasonably detectable.
The original rule without punctuation is not mine, although I have changed the
counts and scores. The modifications with punctuation are my modifications, as
are the run lengths chosen. Personally I use a short one with lowish score in
combination with longer ones with crowbar scores. I'm sure a more scientific
method can be used for score assignment. All I can say about these is WFM.
# ----------------------------------------
#OVERALL% SPAM% HAM% S/O RANK SCORE NAME
# match Bayes-poison lists of lowercase words without articles or common
prepositions
# 18% spam, no ham, good
# 14.493 17.9596 0.0051 1.000 1.00 10.00 PT_WORDLIST_15
body PT_WORDLIST_15 /(?:\b(?!
(?:from|that|have|this|were|with)\b)[a-z]{4,12}\s+){15}/
describe PT_WORDLIST_15 string of 30+ random words
score PT_WORDLIST_15 10.0
# 17% spam, no ham, good
# 13.833 17.1411 0.0051 1.000 1.00 10.00 LW_WORDLIST_20
body LW_WORDLIST_20 /(?:\b(?!
(?:from|that|have|this|were|with)\b)[a-z\.\,\-\;]{4,18}\s+){20}/
describe LW_WORDLIST_20 string of 30+ random words
score LW_WORDLIST_20 10.0
# 15.5% spam, no ham, good
# 12.491 15.4796 0.0000 1.000 1.00 10.00 PT_WORDLIST_20
body PT_WORDLIST_20 /(?:\b(?!
(?:from|that|have|this|were|with)\b)[a-z]{4,12}\s+){20}/
describe PT_WORDLIST_20 string of 30+ random words
score PT_WORDLIST_20 10.0
# 15% spam, no ham, good
# 12.094 14.9873 0.0000 1.000 1.00 10.00 LW_WORDLIST_25
body LW_WORDLIST_25 /(?:\b(?!
(?:from|that|have|this|were|with)\b)[a-z\.\,\-\;]{4,18}\s+){25}/
describe LW_WORDLIST_25 string of 30+ random words
score LW_WORDLIST_25 10.0
# Pretty good, no ham, lotsa spam
# 10.315 12.7829 0.0000 1.000 0.99 10.00 LW_WORDLIST_30
body LW_WORDLIST_30 /(?:\b(?!
(?:from|that|have|this|were|with)\b)[a-z\.\,\-\;]{4,18}\s+){30}/
describe LW_WORDLIST_30 string of 30+ random words
score LW_WORDLIST_30 10.0
# 12% spam, no ham, good
# 10.223 12.6685 0.0000 1.000 0.99 10.00 PT_WORDLIST_25
body PT_WORDLIST_25 /(?:\b(?!
(?:from|that|have|this|were|with)\b)[a-z]{4,12}\s+){25}/
describe PT_WORDLIST_25 string of 30+ random words
score PT_WORDLIST_25 10.0
# Fair amount spam, no ham, but my mod gets more spam, no ham
# 8.172 10.1269 0.0000 1.000 0.99 10.00 PT_WORDLIST_30
body PT_WORDLIST_30 /(?:\b(?!
(?:from|that|have|this|were|with)\b)[a-z]{4,12}\s+){30}/
describe PT_WORDLIST_30 string of 30+ random words
score PT_WORDLIST_30 10.0
------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.
