Robert Menschel <Robert <at> Menschel.net> writes: > > RM> CTS_CONFIDENTIAL has problems, as does CTS_PROPOSITION. CTS_ANONYMOUS is > RM> questionable, but might be useful -- sample of hits is small. > > RM> CTS_APOLOGY hits only ham here. CTS_VIRUSWARN2 hit only ham, someone's > RM> response to a person fooled by the teddy bear virus hoax. (The original > RM> hoaxee's email is not in my corpus, but the response is.)
Bob, Thanks a lot for the run against the corpus. Eventually we'll figure out how to do that on our own email. The ones you noted that have some issues are intended to be part of a meta-rule and I see that the meta rules did OK as far as FPs. They were prompted by spam we are getting in some numbers here, but they might not be real useful in hitting very much spam elsewhere. (maybe not even here in the long run.) Al
