On Apr 05 2006, John Levine wrote:
if he asked
for it and hasn't asked for it to stop, it's legit.
What if he likes to read the list messages on a single topic, but
doesn't want those messages that deal with another topic?
Then he should change his subscription, but that doesn't make the mail
spam.
Or take the ASRG list itself. For argument's sake, ...
Same answer, deal with it any way you want, but mail you asked for
isn't spam. I agree that if you signed up for one thing and they send
you something completely different, which sometimes happens, that's
bad, but splitting hairs isn't going to get anyone anywhere.
Agreed, I'm not looking to impose a single definition of spam, just
point out that consent isn't a panacea.
And my claim is that the act of measuring the operation of the
spamfilter need not affect spammer behaviour directly.
That's fine if you don't want to test filters that reject, greylist,
delay, or otherwise do anything that senders detect. For the other
99% of spam filters, you've got a problem.
I still don't agree, but I'm happy to label the topic as "non-trivial".
If you can obtain FP and FN numbers, then I believe you have a
valuable summary of accuracy for that system, but it won't tell you
about other aspects such as cpu load etc.
Nor will it let you compare that system to others since the FP and FN
numbers you get on different inputs are not necessarily comparable.
That's a minor issue. The basis of filter evaluation is that you analyze
performance in the past to judge performance in the future. If you
truly believe that your own email stream in the near future is
incomparable to your email in the recent past, then you can just pick
all your filters by flipping a coin.
Email streams of different people can well be incomparable, of course,
as can mail of one individual at different periods in time, etc. Even
that has a limited variation however. For example, the TREC 2005 tests
showed a fair amount of consistency over several test corpora. The
variation is not so great as to make FP and FN comparisons on
different corpora meaningless.
--
Laird Breyer.
_______________________________________________
Asrg mailing list
Asrg(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/asrg