You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@spamassassin.apache.org by qu...@apache.org on 2004/06/14 11:40:00 UTC
svn commit: rev 21195 - incubator/spamassassin/trunk/rules
Author: quinlan
Date: Mon Jun 14 02:40:00 2004
New Revision: 21195
Modified:
incubator/spamassassin/trunk/rules/50_scores.cf
Log:
documentation update, remove ancient statistics
Modified: incubator/spamassassin/trunk/rules/50_scores.cf
==============================================================================
--- incubator/spamassassin/trunk/rules/50_scores.cf (original)
+++ incubator/spamassassin/trunk/rules/50_scores.cf Mon Jun 14 02:40:00 2004
@@ -25,77 +25,10 @@
# not assigned here, the default score will be set to 1.
# The following block of scores were generating using the mass-checking
-# scripts, and a genetic algorithm, to determine the optimum scores which
-# resulted in minimum false positives or negatives. The default evolution
-# parameters are very strongly weighted against false positives.
-
-# Set 0
-# SUMMARY for threshold 5.0:
-# Correctly non-spam: 188867 34.75% (99.84% of non-spam corpus)
-# Correctly spam: 333277 61.32% (94.07% of spam corpus)
-# False positives: 307 0.06% (0.16% of nonspam, 66695 weighted)
-# False negatives: 21022 3.87% (5.93% of spam, 69566 weighted)
-# Average score for spam: 17.6 nonspam: 0.3
-# Average for false-pos: 5.8 false-neg: 3.3
-# TOTAL: 543473 100.00%
-# Set 0 Validation
-# SUMMARY for threshold 5.0:
-# Correctly non-spam: 20974 34.73% (99.79% of non-spam corpus)
-# Correctly spam: 37076 61.40% (94.18% of spam corpus)
-# False positives: 45 0.07% (0.21% of nonspam, 5229 weighted)
-# False negatives: 2290 3.79% (5.82% of spam, 7545 weighted)
-# TCR: 15.652485 SpamRecall: 94.183% SpamPrec: 99.879% FP: 0.07% FN: 3.79%
-
-# Set 1
-# SUMMARY for threshold 5.0:
-# Correctly non-spam: 151197 30.53% (99.92% of non-spam corpus)
-# Correctly spam: 335645 67.77% (97.59% of spam corpus)
-# False positives: 115 0.02% (0.08% of nonspam, 32441 weighted)
-# False negatives: 8303 1.68% (2.41% of spam, 27884 weighted)
-# Average score for spam: 24.5 nonspam: 0.3
-# Average for false-pos: 6.2 false-neg: 3.4
-# TOTAL: 495260 100.00%
-# Set 1 Validation
-# SUMMARY for threshold 5.0:
-# Correctly non-spam: 16784 30.50% (99.83% of non-spam corpus)
-# Correctly spam: 37349 67.87% (97.73% of spam corpus)
-# False positives: 28 0.05% (0.17% of nonspam, 3733 weighted)
-# False negatives: 867 1.58% (2.27% of spam, 2904 weighted)
-# TCR: 37.950348 SpamRecall: 97.731% SpamPrec: 99.925% FP: 0.05% FN: 1.58%
-
-# Set 2
-# SUMMARY for threshold 5.0:
-# Correctly non-spam: 188305 53.52% (99.91% of non-spam corpus)
-# Correctly spam: 158249 44.98% (96.87% of spam corpus)
-# False positives: 165 0.05% (0.09% of nonspam, 17679 weighted)
-# False negatives: 5112 1.45% (3.13% of spam, 14605 weighted)
-# Average score for spam: 19.7 nonspam: -4.3
-# Average for false-pos: 6.2 false-neg: 2.9
-# TOTAL: 351831 100.00%
-# Set 2 Validation
-# SUMMARY for threshold 5.0:
-# Correctly non-spam: 20916 53.50% (99.88% of non-spam corpus)
-# Correctly spam: 17562 44.92% (96.75% of spam corpus)
-# False positives: 25 0.06% (0.12% of nonspam, 1392 weighted)
-# False negatives: 589 1.51% (3.25% of spam, 1679 weighted)
-# TCR: 25.421569 SpamRecall: 96.755% SpamPrec: 99.858% FP: 0.06% FN: 1.51%
-
-# Set 3
-# SUMMARY for threshold 5.0:
-# Correctly non-spam: 140024 46.61% (99.96% of non-spam corpus)
-# Correctly spam: 158916 52.90% (99.13% of spam corpus)
-# False positives: 63 0.02% (0.04% of nonspam, 8171 weighted)
-# False negatives: 1400 0.47% (0.87% of spam, 4008 weighted)
-# Average score for spam: 26.7 nonspam: -4.6
-# Average for false-pos: 5.7 false-neg: 2.9
-# TOTAL: 300403 100.00%
-# Set 3 Validation
-# SUMMARY for threshold 5.0:
-# Correctly non-spam: 15550 46.59% (99.90% of non-spam corpus)
-# Correctly spam: 17648 52.87% (99.08% of spam corpus)
-# False positives: 15 0.04% (0.10% of nonspam, 1133 weighted)
-# False negatives: 164 0.49% (0.92% of spam, 437 weighted)
-# TCR: 74.527197 SpamRecall: 99.079% SpamPrec: 99.915% FP: 0.04% FN: 0.49%
+# scripts, and a perceptron to determine the optimum scores which
+# resulted in minimum false positives or negatives. The scores are
+# weighted to produce roughly 1 false positive in 1000 non-spam messages
+# using the default threshold of 5.0.
# Start of generated scores