At 04:12 PM 8/12/2003, Justin Mason wrote:
Yakov Shafranovich writes:
> We do not need to make standards for blocklists, however a small BCP might
> be useful with things like a need to state clearly what your
> listing/delisting policy is, etc. Most of the BCPs for DNSRBLs would
center
> around the need to state what is actually going on in inside one, not
> actually trying to enforce specific policies. Additionally, support for
> consent systems and associated protocols might be useful as well.
BTW is there a BCP for measuring spam filter effectiveness -- e.g. stating
that effectiveness should be measured in terms of portable numbers like
FP% = (false positives as a percentage of nonspam mail)
FN% = (false negatives as a percentage of spam mail)
? This would be valuable.
I've found that often, DNSBL effectiveness is stated as "it blocked this
much spam" with no discussion of false positive or negative rates. Once
the FP% and FN% are measured, the picture becomes a lot clearer ;)
This can be expended to measuring the effectiveness of any anti-spam tool
(within the consent framework this would apply to source tracking
components and consent systems themselves).
If you or anyone else is willing to work on this, go right ahead, just let
the chair know.
Yakov
_______________________________________________
Asrg mailing list
Asrg(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/asrg