> Would some investigations on the true origin of these domains
> help improving the rules or is there just too much spam and nearly
> no other mail coming from these ISPs (most of them located in south
> america)?
In the case of our mass-checks, these rules got quite high scores because
they hit spam and very little or no ham. So an investigation per se might
not be useful, unless it showed that some given domain was indeed completely
bogus.
What could be useful is the ability to mass-check against someone else's
corpus of ham and spam. This could show people, which might be
generalizable to countries or languages, where the rules are less useful, or
not useful at all.
Of course, even if some place is found where the rule is completely
backwards, it still seems to be a good rule in much of America and probably
much of the world. So the rule wouldn't necessarily be removed, but it
could be relegated to a language or country specific set of rules.
Loren
