Daryl -- remind me again -- is this using the GA or the perceptron? it probably should use the GA, judging by the number of oddly 0-scored rules.
Also, any chance you could move it to a subdir of masses, instead of in your sandbox? it'd probably be a more appropriate location... --j. [EMAIL PROTECTED] writes: > Author: dos > Date: Sun Sep 9 16:57:40 2007 > New Revision: 574106 > > URL: http://svn.apache.org/viewvc?rev=574106&view=rev > Log: > updated scores for revision 573961 active rules added since last mass-check > > Modified: > spamassassin/rules/trunk/sandbox/dos/new-rule-score-gen/scores > spamassassin/rules/trunk/sandbox/dos/new-rule-score-gen/scores-set0 > spamassassin/rules/trunk/sandbox/dos/new-rule-score-gen/scores-set1 > spamassassin/rules/trunk/sandbox/dos/new-rule-score-gen/stats-set0 > spamassassin/rules/trunk/sandbox/dos/new-rule-score-gen/stats-set1 > > Modified: spamassassin/rules/trunk/sandbox/dos/new-rule-score-gen/scores > URL: > http://svn.apache.org/viewvc/spamassassin/rules/trunk/sandbox/dos/new-rule-score-gen/scores?rev=574106&r1=574105&r2=574106&view=diff > ============================================================================== > --- spamassassin/rules/trunk/sandbox/dos/new-rule-score-gen/scores (original) > +++ spamassassin/rules/trunk/sandbox/dos/new-rule-score-gen/scores Sun Sep 9 > 16:57:40 2007 > @@ -1,32 +1,40 @@ > -score AXB_MIME_IMG830 0.000 2.799 0.000 0.000 > -score AXB_RCVD_ZOOBSEND 0.000 2.835 0.000 0.000 > -score AXB_RCVD_ZOONAT 0.000 1.000 0.000 0.000 > -score AXB_XTIDX_CHAIN 0.000 1.123 0.000 0.000 > -score CARD_DIRECT_WWW_ADDRESS 3.185 1.121 0.000 0.000 > -score DOS_OE_TO_MX 0.000 2.732 0.000 0.000 > -score DOS_OUTLOOK_TO_MX 3.199 2.658 0.000 0.000 > -score DOS_OUTLOOK_TO_MX_IMAGE 0.000 1.000 0.000 0.000 > -score DOS_STOCK_CDYV_GENERIC 0.000 0.000 0.000 0.000 > -score FB_ACHIEVE_BACH 1.000 1.027 0.000 0.000 > -score FB_B0NUS 0.000 1.000 0.000 0.000 > -score FB_MED1CAT 0.000 0.630 0.000 0.000 > -score FB_SAVE_PERSC 3.599 1.000 0.000 0.000 > -score FB_SMALL_PEN 0.000 1.072 0.000 0.000 > -score FB_STRONGER_EJ 1.000 0.527 0.000 0.000 > -score FB_WITHOUT_PRESC 0.000 1.220 0.000 0.000 > -score FH_FAKE_RCVD_LINE_B 1.250 2.199 0.000 0.000 > -score FM_MORTGAGE3PLUS 0.000 1.615 0.000 0.000 > -score FRT_BEFORE 3.503 2.770 0.000 0.000 > -score FRT_OPPORTUN1 0.000 2.475 0.000 0.000 > -score JM_SOUGHT_2 4.499 4.275 0.000 0.000 > -score JM_SOUGHT_3 4.399 3.999 0.000 0.000 > -score KAM_LOTTO1 0.000 3.099 0.000 0.000 > -score KAM_LOTTO2 0.000 1.000 0.000 0.000 > -score KAM_LOTTO3 0.000 1.009 0.000 0.000 > -score LOTTERY_PH_004470 2.999 2.651 0.000 0.000 > -score STOX_RCVD_N_NN_N 0.000 1.000 0.000 0.000 > -score TVD_PDF_FINGER01 0.000 1.129 0.000 0.000 > -score TVD_PDF_FINGER01_JO 3.991 1.521 0.000 0.000 > +score AB_TEST_PDF1 0.581 0.000 0.000 0.000 > +score AXB_RCVD_ZOOBSEND 4.099 2.782 0.000 0.000 > +score AXB_RCVD_ZOONAT 2.431 2.359 0.000 0.000 > +score AXB_XTIDX_CHAIN 0.447 3.778 0.000 0.000 > +score BC_ENCODED_PDF 0.491 0.000 0.000 0.000 > +score CARD_DIRECT_WWW_ADDRESS 2.441 0.931 0.000 0.000 > +score DOS_OE_TO_MX 1.942 0.720 0.000 0.000 > +score DOS_OUTLOOK_TO_MX 3.034 1.009 0.000 0.000 > +score DOS_OUTLOOK_TO_MX_IMAGE 1.524 3.919 0.000 0.000 > +score FB_ACHIEVE_BACH 1.000 1.000 0.000 0.000 > +score FB_B0NUS 1.243 0.816 0.000 0.000 > +score FB_CASINO 0.733 1.671 0.000 0.000 > +score FB_MED1CAT 1.000 1.761 0.000 0.000 > +score FB_SAVE_PERSC 2.709 1.071 0.000 0.000 > +score FB_SMALL_PEN 0.394 0.593 0.000 0.000 > +score FB_STRONGER_EJ 1.000 1.000 0.000 0.000 > +score FB_WITHOUT_PRESC 2.331 1.067 0.000 0.000 > +score FH_FAKE_RCVD_LINE_B 0.943 2.187 0.000 0.000 > +score FM_MORTGAGE3PLUS 0.224 0.001 0.000 0.000 > +score FRT_BEFORE 1.000 3.136 0.000 0.000 > +score FRT_ERECTION 1.026 0.367 0.000 0.000 > +score FRT_OPPORTUN1 1.657 2.978 0.000 0.000 > +score FS_ABIGGER 2.279 1.000 0.000 0.000 > +score FS_LOWER_YOUR 1.653 0.000 0.000 0.000 > +score FS_LOW_INTMOR 1.000 0.000 0.000 0.000 > +score FS_WEIGHT_LOSS 2.922 1.000 0.000 0.000 > +score JM_SOUGHT_2 3.216 3.912 0.000 0.000 > +score JM_SOUGHT_3 4.499 3.833 0.000 0.000 > +score KAM_LOTTO1 0.873 3.043 0.000 0.000 > +score KAM_LOTTO2 1.526 1.197 0.000 0.000 > +score KAM_LOTTO3 1.000 2.039 0.000 0.000 > +score LOTTERY_PH_004470 1.000 1.991 0.000 0.000 > +score STOX_META_5 0.001 0.753 0.000 0.000 > +score STOX_RCVD_N_NN_N 1.924 1.349 0.000 0.000 > +score TVD_ENHANCE 0.340 0.000 0.000 0.000 > +score TVD_PDF_FINGER01 1.000 1.346 0.000 0.000 > +score TVD_PDF_FINGER01_JO 1.371 1.514 0.000 0.000 > score URIBL_L1SPEWS 0.000 0.000 0.000 0.000 > score URIBL_L2SPEWS 0.000 0.000 0.000 0.000 > score WHOIS_MONIKER_ROLE 0.000 0.000 0.000 0.000 > > Modified: spamassassin/rules/trunk/sandbox/dos/new-rule-score-gen/scores-set0 > URL: > http://svn.apache.org/viewvc/spamassassin/rules/trunk/sandbox/dos/new-rule-score-gen/scores-set0?rev=574106&r1=574105&r2=574106&view=diff > ============================================================================== > --- spamassassin/rules/trunk/sandbox/dos/new-rule-score-gen/scores-set0 > (original) > +++ spamassassin/rules/trunk/sandbox/dos/new-rule-score-gen/scores-set0 Sun > Sep 9 16:57:40 2007 > @@ -1,19 +1,43 @@ > -# Using score set 0 logs for revision 573507 from: > -# ham-bb-fredt.log ham-bb-jm.log ham-bb-zmi.log ham-daf.log ham-dos.log > ham-jm.log ham-theo.log spam-bb-fredt.log spam-bb-jm.log spam-bb-zmi.log > spam-daf.log spam-dos.log spam-jm.log spam-theo.log > +# Using score set 0 logs for revision 573961 from: > +# ham-bb-fredt.log ham-bb-jm.log ham-bb-zmi.log ham-daf.log ham-theo.log > spam-bb-fredt.log spam-bb-jm.log spam-bb-zmi.log spam-daf.log spam-theo.log > > -score CARD_DIRECT_WWW_ADDRESS 3.185 > -score DOS_OUTLOOK_TO_MX 3.199 > +score AB_TEST_PDF1 0.581 > +score AXB_RCVD_ZOOBSEND 4.099 > +score AXB_RCVD_ZOONAT 2.431 > +score AXB_XTIDX_CHAIN 0.447 > +score BC_ENCODED_PDF 0.491 > +score CARD_DIRECT_WWW_ADDRESS 2.441 > +score DOS_OE_TO_MX 1.942 > +score DOS_OUTLOOK_TO_MX 3.034 > +score DOS_OUTLOOK_TO_MX_IMAGE 1.524 > score FB_ACHIEVE_BACH 1.000 > -score FB_CASINO 1.392 > -score FB_SAVE_PERSC 3.599 > +score FB_B0NUS 1.243 > +score FB_CASINO 0.733 > +score FB_MED1CAT 1.000 > +score FB_SAVE_PERSC 2.709 > +score FB_SMALL_PEN 0.394 > score FB_STRONGER_EJ 1.000 > -score FH_FAKE_RCVD_LINE_B 1.250 > -score FRT_BEFORE 3.503 > -score JM_SOUGHT_2 4.499 > -score JM_SOUGHT_3 4.399 > -score LOTTERY_PH_004470 2.999 > -score STOX_META_5 0.639 > -score TVD_PDF_FINGER01_JO 3.991 > +score FB_WITHOUT_PRESC 2.331 > +score FH_FAKE_RCVD_LINE_B 0.943 > +score FM_MORTGAGE3PLUS 0.224 > +score FRT_BEFORE 1.000 > +score FRT_ERECTION 1.026 > +score FRT_OPPORTUN1 1.657 > +score FS_ABIGGER 2.279 > +score FS_LOWER_YOUR 1.653 > +score FS_LOW_INTMOR 1.000 > +score FS_WEIGHT_LOSS 2.922 > +score JM_SOUGHT_2 3.216 > +score JM_SOUGHT_3 4.499 > +score KAM_LOTTO1 0.873 > +score KAM_LOTTO2 1.526 > +score KAM_LOTTO3 1.000 > +score LOTTERY_PH_004470 1.000 > +score STOX_META_5 0.001 > +score STOX_RCVD_N_NN_N 1.924 > +score TVD_ENHANCE 0.340 > +score TVD_PDF_FINGER01 1.000 > +score TVD_PDF_FINGER01_JO 1.371 > # in active.list but have no hits in recent corpus > score URIBL_L1SPEWS 0.000 > score URIBL_L2SPEWS 0.000 > > Modified: spamassassin/rules/trunk/sandbox/dos/new-rule-score-gen/scores-set1 > URL: > http://svn.apache.org/viewvc/spamassassin/rules/trunk/sandbox/dos/new-rule-score-gen/scores-set1?rev=574106&r1=574105&r2=574106&view=diff > ============================================================================== > --- spamassassin/rules/trunk/sandbox/dos/new-rule-score-gen/scores-set1 > (original) > +++ spamassassin/rules/trunk/sandbox/dos/new-rule-score-gen/scores-set1 Sun > Sep 9 16:57:40 2007 > @@ -1,34 +1,38 @@ > # Using score set 1 logs for revision 573802 from: > -# ham-net-daf.log ham-net-dos.log ham-net-jm.log spam-net-daf.log > spam-net-dos.log spam-net-jm.log > +# ham-net-daf.log ham-net-dos.log ham-net-jm.log ham-net-theo.log > spam-net-daf.log spam-net-dos.log spam-net-jm.log spam-net-theo.log > > -score AXB_MIME_IMG830 2.799 > -score AXB_RCVD_ZOOBSEND 2.835 > -score AXB_RCVD_ZOONAT 1.000 > -score AXB_XTIDX_CHAIN 1.123 > -score CARD_DIRECT_WWW_ADDRESS 1.121 > -score DOS_OE_TO_MX 2.732 > -score DOS_OUTLOOK_TO_MX 2.658 > -score DOS_OUTLOOK_TO_MX_IMAGE 1.000 > -score FB_ACHIEVE_BACH 1.027 > -score FB_B0NUS 1.000 > -score FB_MED1CAT 0.630 > -score FB_SAVE_PERSC 1.000 > -score FB_SMALL_PEN 1.072 > -score FB_STRONGER_EJ 0.527 > -score FB_WITHOUT_PRESC 1.220 > -score FH_FAKE_RCVD_LINE_B 2.199 > -score FM_MORTGAGE3PLUS 1.615 > -score FRT_BEFORE 2.770 > -score FRT_OPPORTUN1 2.475 > -score JM_SOUGHT_2 4.275 > -score JM_SOUGHT_3 3.999 > -score KAM_LOTTO1 3.099 > -score KAM_LOTTO2 1.000 > -score KAM_LOTTO3 1.009 > -score LOTTERY_PH_004470 2.651 > -score STOX_RCVD_N_NN_N 1.000 > -score TVD_PDF_FINGER01 1.129 > -score TVD_PDF_FINGER01_JO 1.521 > +score AXB_RCVD_ZOOBSEND 2.782 > +score AXB_RCVD_ZOONAT 2.359 > +score AXB_XTIDX_CHAIN 3.778 > +score CARD_DIRECT_WWW_ADDRESS 0.931 > +score DOS_OE_TO_MX 0.720 > +score DOS_OUTLOOK_TO_MX 1.009 > +score DOS_OUTLOOK_TO_MX_IMAGE 3.919 > +score FB_ACHIEVE_BACH 1.000 > +score FB_B0NUS 0.816 > +score FB_CASINO 1.671 > +score FB_MED1CAT 1.761 > +score FB_SAVE_PERSC 1.071 > +score FB_SMALL_PEN 0.593 > +score FB_STRONGER_EJ 1.000 > +score FB_WITHOUT_PRESC 1.067 > +score FH_FAKE_RCVD_LINE_B 2.187 > +score FM_MORTGAGE3PLUS 0.001 > +score FRT_BEFORE 3.136 > +score FRT_ERECTION 0.367 > +score FRT_OPPORTUN1 2.978 > +score FS_ABIGGER 1.000 > +score FS_WEIGHT_LOSS 1.000 > +score JM_SOUGHT_2 3.912 > +score JM_SOUGHT_3 3.833 > +score KAM_LOTTO1 3.043 > +score KAM_LOTTO2 1.197 > +score KAM_LOTTO3 2.039 > +score LOTTERY_PH_004470 1.991 > +score STOX_META_5 0.753 > +score STOX_RCVD_N_NN_N 1.349 > +score TVD_PDF_FINGER01 1.346 > +score TVD_PDF_FINGER01_JO 1.514 > # in active.list but have no hits in recent corpus > score DOS_STOCK_CDYV_GENERIC 0.000 > score URIBL_L1SPEWS 0.000 > > Modified: spamassassin/rules/trunk/sandbox/dos/new-rule-score-gen/stats-set0 > URL: > http://svn.apache.org/viewvc/spamassassin/rules/trunk/sandbox/dos/new-rule-score-gen/stats-set0?rev=574106&r1=574105&r2=574106&view=diff > ============================================================================== > --- spamassassin/rules/trunk/sandbox/dos/new-rule-score-gen/stats-set0 > (original) > +++ spamassassin/rules/trunk/sandbox/dos/new-rule-score-gen/stats-set0 Sun > Sep 9 16:57:40 2007 > @@ -1,40 +1,40 @@ > ##### WITH NEW RULES AND SCORES ##### > > # SUMMARY for threshold 5.0: > -# Correctly non-spam: 88673 15.058% (96.738% of non-spam corpus) > -# Correctly spam: 425517 72.260% (85.582% of spam corpus) > -# False positives: 2990 0.508% (3.262% of nonspam, 204215 weighted) > -# False negatives: 71689 12.174% (14.418% of spam, 175874 weighted) > -# Average score for spam: 13.5 nonspam: 1.1 > -# Average for false-pos: 6.3 false-neg: 2.5 > -# TOTAL: 588869 100.00% > +# Correctly non-spam: 76443 16.359% (96.129% of non-spam corpus) > +# Correctly spam: 329794 70.577% (85.051% of spam corpus) > +# False positives: 3078 0.659% (3.871% of nonspam, 189994 weighted) > +# False negatives: 57967 12.405% (14.949% of spam, 138775 weighted) > +# Average score for spam: 14.0 nonspam: 1.2 > +# Average for false-pos: 6.3 false-neg: 2.4 > +# TOTAL: 467282 100.00% > > Reading scores from "tmprules"... > Reading per-message hit stat logs and scores... > > # SUMMARY for threshold 5.0: > -# Correctly non-spam: 11216 96.83% > -# Correctly spam: 53135 85.61% > -# False positives: 367 3.17% > -# False negatives: 8934 14.39% > -# TCR(l=50): 2.274923 SpamRecall: 85.606% SpamPrec: 99.314% > +# Correctly non-spam: 9653 96.15% > +# Correctly spam: 41058 84.98% > +# False positives: 387 3.85% > +# False negatives: 7256 15.02% > +# TCR(l=50): 1.815906 SpamRecall: 84.982% SpamPrec: 99.066% > > ##### WITHOUT NEW RULES AND SCORES ##### > Reading scores from "../rules-base"... > Reading per-message hit stat logs and scores... > > # SUMMARY for threshold 5.0: > -# Correctly non-spam: 88932 97.02% > -# Correctly spam: 239195 48.11% > -# False positives: 2731 2.98% > -# False negatives: 258011 51.89% > -# TCR(l=50): 1.260150 SpamRecall: 48.108% SpamPrec: 98.871% > +# Correctly non-spam: 76702 96.46% > +# Correctly spam: 203227 52.41% > +# False positives: 2819 3.54% > +# False negatives: 184534 47.59% > +# TCR(l=50): 1.191337 SpamRecall: 52.410% SpamPrec: 98.632% > Reading scores from "../rules-base"... > Reading per-message hit stat logs and scores... > > # SUMMARY for threshold 5.0: > -# Correctly non-spam: 11247 97.10% > -# Correctly spam: 29967 48.28% > -# False positives: 336 2.90% > -# False negatives: 32102 51.72% > -# TCR(l=50): 1.269253 SpamRecall: 48.280% SpamPrec: 98.891% > +# Correctly non-spam: 9683 96.44% > +# Correctly spam: 25357 52.48% > +# False positives: 357 3.56% > +# False negatives: 22957 47.52% > +# TCR(l=50): 1.183964 SpamRecall: 52.484% SpamPrec: 98.612% > > Modified: spamassassin/rules/trunk/sandbox/dos/new-rule-score-gen/stats-set1 > URL: > http://svn.apache.org/viewvc/spamassassin/rules/trunk/sandbox/dos/new-rule-score-gen/stats-set1?rev=574106&r1=574105&r2=574106&view=diff > ============================================================================== > --- spamassassin/rules/trunk/sandbox/dos/new-rule-score-gen/stats-set1 > (original) > +++ spamassassin/rules/trunk/sandbox/dos/new-rule-score-gen/stats-set1 Sun > Sep 9 16:57:40 2007 > @@ -1,40 +1,40 @@ > ##### WITH NEW RULES AND SCORES ##### > > # SUMMARY for threshold 5.0: > -# Correctly non-spam: 13407 9.978% (99.777% of non-spam corpus) > -# Correctly spam: 118413 88.130% (97.923% of spam corpus) > -# False positives: 30 0.022% (0.223% of nonspam, 8462 weighted) > -# False negatives: 2512 1.870% (2.077% of spam, 5494 weighted) > -# Average score for spam: 26.9 nonspam: -0.9 > -# Average for false-pos: 6.3 false-neg: 2.2 > -# TOTAL: 134362 100.00% > +# Correctly non-spam: 31436 10.540% (98.924% of non-spam corpus) > +# Correctly spam: 262597 88.044% (98.543% of spam corpus) > +# False positives: 342 0.115% (1.076% of nonspam, 86383 weighted) > +# False negatives: 3883 1.302% (1.457% of spam, 10106 weighted) > +# Average score for spam: 25.8 nonspam: 0.2 > +# Average for false-pos: 6.0 false-neg: 2.6 > +# TOTAL: 298258 100.00% > > Reading scores from "tmprules"... > Reading per-message hit stat logs and scores... > > # SUMMARY for threshold 5.0: > -# Correctly non-spam: 1673 99.94% > -# Correctly spam: 14880 97.96% > -# False positives: 1 0.06% > -# False negatives: 310 2.04% > -# TCR(l=50): 42.194444 SpamRecall: 97.959% SpamPrec: 99.993% > +# Correctly non-spam: 3978 98.96% > +# Correctly spam: 32844 98.53% > +# False positives: 42 1.04% > +# False negatives: 489 1.47% > +# TCR(l=50): 12.874855 SpamRecall: 98.533% SpamPrec: 99.872% > > ##### WITHOUT NEW RULES AND SCORES ##### > Reading scores from "../rules-base"... > Reading per-message hit stat logs and scores... > > # SUMMARY for threshold 5.0: > -# Correctly non-spam: 13420 99.87% > -# Correctly spam: 115207 95.27% > -# False positives: 17 0.13% > -# False negatives: 5718 4.73% > -# TCR(l=50): 18.411236 SpamRecall: 95.271% SpamPrec: 99.985% > +# Correctly non-spam: 31293 98.47% > +# Correctly spam: 255766 95.98% > +# False positives: 485 1.53% > +# False negatives: 10714 4.02% > +# TCR(l=50): 7.621554 SpamRecall: 95.979% SpamPrec: 99.811% > Reading scores from "../rules-base"... > Reading per-message hit stat logs and scores... > > # SUMMARY for threshold 5.0: > -# Correctly non-spam: 1674 100.00% > -# Correctly spam: 14508 95.51% > -# False positives: 0 0.00% > -# False negatives: 682 4.49% > -# TCR(l=50): 22.272727 SpamRecall: 95.510% SpamPrec: 100.000% > +# Correctly non-spam: 3963 98.58% > +# Correctly spam: 32042 96.13% > +# False positives: 57 1.42% > +# False negatives: 1291 3.87% > +# TCR(l=50): 8.049505 SpamRecall: 96.127% SpamPrec: 99.822%
