recently I get a lot of “postcard” spam. It is always structured the same way,
but they change to who it is (friends, colleage, family, etc.), and from
which postcard service. I have trained spamsieve with a couple of them,
but it still doesn’t catch them. I could train more, and I presume it would
catch them, but then I am worried that legitimate emails containing
similar sets of words won’t come through.
Given however that the phrasing of these messages is rather distinct,
can spamsieve recognize phrases?
Perhaps that’s because of the other spam messages that you let SpamSieve think were good.
I would’t worry about it. When properly trained with both your good messages and your spam ones, SpamSieve is extremely accurate and is particularly good at avoiding false positives (good messages classified as spam).
Ah, yes, that of course makes sense. So, I tried to mark the others.
Well, I don’t have too many examples of good ecards. But then, it’s not too
tragic if I miss those.
Thank you for the reply.