On 27/Jan/10 16:48, Andrew Richards wrote:
Surely there's no real problem if someone mis-classifies a message as spam
or non-spam:
However, the possibility of retracting a (mis-)classification may be
useful in some cases.
If I'm providing a mail service I would like to aggregate
results so that when the same message is classified as spam several times by
different users I'll start blocking similar messages
Hm... it is not straightforward to establish whether new messages are
/similar/ to already received ones. Checking an authenticated id, e.g.
for messages received "with ESMTP[S]A", would be easier.
- either doing this aggregation myself if I'm a big ISP etc.,
or using an external service providing this functionality.
I'd be interested in the external service, not only because I'm not a
big ISP, but also for the sake of neutrality, transparency, etc.
Obviously, it has to receive abuse reports... Perhaps not all the
world's ones, just some. Which ones, by sending domain?
Is this service a Mythical Global Reputation System (TMGRS)?
_______________________________________________
Asrg mailing list
Asrg(_at_)irtf(_dot_)org
http://www.irtf.org/mailman/listinfo/asrg