Back in http://mail.python.org/pipermail/spambayes/2006-January/018702.html , Tim Peters and I had a dialog about training on unusual ham - monthly messages from http://www.boldtype.com. I just got another one and it scored 50% on the spam scale. The clues follow - I'd really appreciate any help. Thanks, Bob
Combined Score: 50% (0.5) Internal ham score (*H*): 1 Internal spam score (*S*): 1 # ham trained on: 1229 # spam trained on: 20331 150 Significant Tokens token spamprob #ham #spam 'policy:' 0.00858658 32 1 'url:dots' 0.0115681 19 0 'url:books' 0.0121951 18 0 'unsubscribe:' 0.0129932 56 9 'url:gif)' 0.0136778 16 0 'url:covers' 0.0180723 12 0 'respects' 0.0186053 32 7 'url:store' 0.019152 28 6 'url:ad' 0.0192872 19 3 'criticism,' 0.0196507 11 0 'pages:' 0.0196507 11 0 'url:func' 0.0196507 11 0 'url:grey_dot' 0.0196507 11 0 'c/o' 0.0206833 15 2 'list.' 0.0207364 51 15 'list,' 0.0211427 89 29 '10012' 0.0215311 10 0 '1212' 0.0215311 10 0 '594' 0.0215311 10 0 'anjuli' 0.0215311 10 0 'ayer' 0.0215311 10 0 'boldtype' 0.0215311 10 0 'boldtype,' 0.0215311 10 0 'boldtype.com' 0.0215311 10 0 'complies' 0.0215311 10 0 'email addr:boldtype.com' 0.0215311 10 0 'email addr:boldtype.com.' 0.0215311 10 0 'email name:subscriptions' 0.0215311 10 0 'email-based' 0.0215311 10 0 'flavorpill' 0.0215311 10 0 'from:addr:boldtype.com' 0.0215311 10 0 'from:name:boldtype' 0.0215311 10 0 'glei' 0.0215311 10 0 'graphical' 0.0215311 10 0 'handpicked' 0.0215311 10 0 'laster' 0.0215311 10 0 'magazines:' 0.0215311 10 0 'mangan' 0.0215311 10 0 'monthly,' 0.0215311 10 0 'npr' 0.0215311 10 0 'publishers,' 0.0215311 10 0 'sascha' 0.0215311 10 0 'subject: | ' 0.0215311 10 0 'submissions' 0.0215311 10 0 'url:108s877' 0.0215311 10 0 'url:392388' 0.0215311 10 0 'url:boldtype' 0.0215311 10 0 'url:boldtype_com' 0.0215311 10 0 'url:boldtype_logo' 0.0215311 10 0 'url:border-pixel' 0.0215311 10 0 'url:current' 0.0215311 10 0 'url:designby' 0.0215311 10 0 'url:earplug' 0.0215311 10 0 'url:editor' 0.0215311 10 0 'url:email_address' 0.0215311 10 0 'url:email_subject' 0.0215311 10 0 'url:flavorpill' 0.0215311 10 0 'url:issues' 0.0215311 10 0 'url:jcreport' 0.0215311 10 0 'url:job_id' 0.0215311 10 0 'url:listing_id' 0.0215311 10 0 'url:mail_list_id' 0.0215311 10 0 'url:other-pubs_culture' 0.0215311 10 0 'url:other-pubs_earplug' 0.0215311 10 0 'url:other-pubs_fashion' 0.0215311 10 0 'url:other-pubs_flavorpill' 0.0215311 10 0 'url:other-pubs_jcreport' 0.0215311 10 0 'url:other-pubs_music' 0.0215311 10 0 'url:partnerships' 0.0215311 10 0 'url:spamlaws' 0.0215311 10 0 'url:sublit' 0.0215311 10 0 'url:subscriber_id' 0.0215311 10 0 'url:}' 0.0215311 10 0 'weissman' 0.0215311 10 0 'nonfiction' 0.0215731 12 1 'url:monthly' 0.0215731 12 1 "insider's" 0.0220053 14 2 'url:mail' 0.0227833 41 13 'reviews,' 0.0233401 11 1 'url:archive' 0.0233401 11 1 'url:sub' 0.0233401 11 1 'publisher:' 0.0238095 9 0 'url:grey_separation2' 0.0238095 9 0 'url:zoom_in' 0.0238095 9 0 '1950' 0.97619 0 9 'abigail' 0.97619 0 9 'clifford' 0.97619 0 9 'handkerchief' 0.97619 0 9 'image.' 0.97619 0 9 'prospect,' 0.97619 0 9 'rogue' 0.97619 0 9 'belle' 0.978469 0 10 'boyle' 0.978469 0 10 'celluloid' 0.978469 0 10 'errol' 0.978469 0 10 'gritty' 0.978469 0 10 'mann' 0.978469 0 10 'overlooked' 0.978469 0 10 'beethoven' 0.980349 0 11 'cecil' 0.980349 0 11 'crafty' 0.980349 0 11 'examine' 0.980349 0 11 'whims' 0.980349 0 11 '1962' 0.981928 0 12 '1983' 0.981928 0 12 'alfred' 0.981928 0 12 'dismissed' 0.981928 0 12 'spanned' 0.981928 0 12 'gentlemen,' 0.983271 0 13 'jorge' 0.983271 0 13 'parish' 0.983271 0 13 'proximity' 0.983271 0 13 'readily' 0.983271 0 13 'chaplin' 0.984429 0 14 'comparison.' 0.984429 0 14 'dreamlike' 0.984429 0 14 'explodes' 0.984429 0 14 'stars,' 0.984429 0 14 'otto' 0.985437 0 15 '1985' 0.987106 0 17 'atom' 0.987106 0 17 'bourgeoisie' 0.987106 0 17 'reject' 0.987106 0 17 'sued' 0.987106 0 17 'temporarily' 0.987106 0 17 'imagined' 0.987805 0 18 'praise' 0.987805 0 18 'emperor' 0.988432 0 19 'syndrome.' 0.988432 0 19 'balfour' 0.988998 0 20 'escaped' 0.988998 0 20 'finalists' 0.988998 0 20 'anxiety,' 0.989978 0 22 'raging' 0.989978 0 22 'tucked' 0.989978 0 22 'apt' 0.990798 0 24 'francis' 0.990798 0 24 'face.' 0.991159 0 25 'marie' 0.992091 0 28 'respectful' 0.99236 0 29 'dose' 0.992611 0 30 'nineteen' 0.993469 0 34 'disliked' 0.994572 0 41 'awesome' 0.99505 0 45 'ego' 0.996397 0 62 'ladies' 0.99676 0 69 'projection' 0.997168 0 79 'broke' 0.997512 0 90 'accordance' 0.998921 0 208 'discreet' 0.999019 0 229 _______________________________________________ [email protected] http://mail.python.org/mailman/listinfo/spambayes Check the FAQ before asking: http://spambayes.sf.net/faq.html
