On Wed, Feb 27, 2002 at 04:28:07PM -0800, Craig R Hughes wrote: > Duncan Findlay wrote: > > > Ummm... I'd be heavily inclined to set these spam scores to 0.01. It's not > > that I don't trust the GA, it's just that if these are the outputs, they > > aren't needed in the first place. > > That's not necessarily the case. They might be needed to reduce false > positives. As I posted a couple messages ago, I think I'll do some amount of > investigation of the tests which are hit by remaining false positives and false > negatives to see which rules are affecting those messages and get a little more > information about what's going on. For the curious, I've attached the frequency > file that the GA used as part of the process of determining these scores. The > format of the file is pretty straightforward. > > C
Questional tests: score GAPPY_TEXT -3.667 spam: 261. nonspam: 112. score PORN_8 -5.452 spam: 3. nonspam: 22. score TRACKER_ID -4.899 spam: 9. nonspam: 9 Ridiculous scores: score 25FREEMEGS_URL -4.606 spam: 6. nonspam: 0. score CYBER_FIRE_POWER -4.020 spam: 10. nonspam: 0. score EXCUSE_5 13.447 spam: 10. nonspam: 0. score MONSTERHUT -8.280 spam: 40. nonspam: 0. score ONCE_IN_LIFETIME -4.604 spam: 83. nonspam: 5. -- Duncan Findlay _______________________________________________ Spamassassin-talk mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/spamassassin-talk
