Bart Schaefer <[EMAIL PROTECTED]> writes: > score: ASCII_FORM_ENTRY 0.036 -> -1.660 > score: BUGZILLA_BUG -2.000 -> 0.921
BUGZILLA_BUG obviously needs to be fixed. Maybe an eval would be best. > score: DATE_MISSING 0.248 -> -2.140 > score: EXCUSE_16 1.345 -> -0.721 > score: FORGED_HOTMAIL_RCVD 0.530 -> -0.356 > score: FROM_AND_TO_SAME 0.877 -> -2.071 > score: FROM_NAME_NO_SPACES 0.500 -> -0.114 > score: GREEN_EXCUSE_1 3.116 -> -2.019 > score: INTL_EXEC_GUILD 0.781 -> -0.039 > score: MONEY_BACK 1.489 -> -0.239 > score: MONEY_MAKING 2.490 -> -0.687 > score: MSGID_CHARS_WEIRD 1.500 -> -2.178 > score: NO_REAL_NAME 0.632 -> -1.068 > score: X_NOT_PRESENT 0.500 -> -1.920 I'm removing X_NOT_PRESENT and MSGID_CHARS_WEIRD from both trees. I think FROM_NAME_NO_SPACES is a good rule that needs more tweaking. Several are probably oddbads because they have low frequency. I have zero hits for DATE_MISSING. Some others look recoverable. Rule spam good ASCII_FORM_ENTRY 42 8 BUGZILLA_BUG 0 178 DATE_MISSING 0 0 EXCUSE_16 102 7 FORGED_HOTMAIL_RCVD 124 2 FROM_AND_TO_SAME 94 125 FROM_NAME_NO_SPACES 471 87 GREEN_EXCUSE_1 1 0 INTL_EXEC_GUILD 0 0 MONEY_BACK 41 0 MONEY_MAKING 19 0 MSGID_CHARS_WEIRD 11 1 NO_REAL_NAME 875 789 X_NOT_PRESENT 601 614 > How did these get exactly 1.0? Not represented in the corpus at all? > > score: FORGED_RCVD_TRAIL absent -> 1.000 > score: FROM_ADDRESS_EQ_REAL absent -> 1.000 > score: TO_ADDRESS_EQ_REAL absent -> 1.000 You're looking at HEAD. These are new rules I added last night. Dan _______________________________________________________________ Don't miss the 2002 Sprint PCS Application Developer's Conference August 25-28 in Las Vegas - http://devcon.sprintpcs.com/adp/index.cfm?source=osdntextlink _______________________________________________ Spamassassin-talk mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/spamassassin-talk
