http://mywebpages.comcast.net/mkettler/sa/antidrug.cf
Sorry for the "release of the day" phenomenon.. That'll teach me to skip some of my regression tests when making an new release..
In 0.50 and 0.51 LOCAL_DRUGS_MALEDYSFUNCTION_OBFU will false-hit on an email containing an un-obfuscated v-pill name.
Fixing this halves the hitrate of the OBFU rule, but it's still much better in hitrate than 0.43 was.
micro-corpus mass-check results:
0.52
OVERALL SPAM HAM NAME
8073 7141 932 (all messages)
1566 1566 0 LOCAL_DRUGS_MALEDYSFUNCTION
717 717 0 LOCAL_DRUGS_DIET
666 666 0 LOCAL_DRUGS_PAIN
634 634 0 LOCAL_DRUGS_DIET_MALEDYS
596 596 0 LOCAL_DRUGS_MANYKINDS
594 594 0 LOCAL_DRUGS_MUSCLE
550 550 0 LOCAL_DRUGS_MALEDYSFUNCTION_OBFU
467 467 0 LOCAL_DRUGS_PAIN_MALEDYS
466 466 0 LOCAL_DRUGS_DIET_PAIN
385 385 0 LOCAL_DRUGS_DEPRESSION
382 382 0 LOCAL_DRUGS_ANXIETY_MALEDYS
345 345 0 LOCAL_DRUGS_DEPRESSION_MALEDYS
532 531 1 LOCAL_DRUGS_ANXIETY
388 387 1 LOCAL_DRUGS_SLEEP
0.43 on same corpus (for comparison/sanity check)
OVERALL SPAM HAM NAME
8073 7141 932 (all messages)
1453 1453 0 LOCAL_DRUGS_MALEDYSFUNCTION
653 653 0 LOCAL_DRUGS_DIET
632 632 0 LOCAL_DRUGS_PAIN
555 555 0 LOCAL_DRUGS_DIET_MALEDYS
498 498 0 LOCAL_DRUGS_MUSCLE
483 483 0 LOCAL_DRUGS_MANYKINDS
437 437 0 LOCAL_DRUGS_DIET_PAIN
422 422 0 LOCAL_DRUGS_PAIN_MALEDYS
316 316 0 LOCAL_DRUGS_DEPRESSION
283 283 0 LOCAL_DRUGS_ANXIETY_MALEDYS
270 270 0 LOCAL_DRUGS_DEPRESSION_MALEDYS
196 196 0 LOCAL_DRUGS_MALDYSFUNCTION_OBFU
427 426 1 LOCAL_DRUGS_ANXIETY
289 288 1 LOCAL_DRUGS_SLEEP