Spyware Warrior

[hornet777]

Perhaps another of hornet's impossibly stupid questions, but I wonder aloud if spammers have the ability to detect what filter(s) one has constructed in a general net-based account and adjust subjects/bodies accordingly? Maybe its just me, but it seems like this is so, for whenever I adjust a filter, it seems as though the spammer adapts.

Not a lot (quantity -- <10/day) of spam, but any at all annoys me, so I try to keep ahead of it.

[Erikalbert]

I can't see how it would work particularly for client based filters particularly those indidvualised bayesian filters.

I suppose if the spammer has some way of knowing whether a user opens the mail or not (via webbugs ??) they might have some feedback on whether it made it through, but even then it would be iffy, because many users delete spam mail without opening it at all just by looking at the headers.

[Oldfrog]

I agree with Ericalbert that there is very little chance of a spammer detecting your individual filter rules. I also believe that the webbugs are more useful for determining that an account is active than for anything else.

One thing that spammers do know is that there are filters available and that more and more people use them all the time. Knowing that, it only makes sense that they (the spammers) would continually change up their content in an effort to slip through.

[thejynxed]

If you use bayesian filtering, alot of spammers now do what is called "bayespoisoning" where they untrain your filters by specially crafting the headers of spam a certain way.

Another thing you might notice, if you use GMail, is that lately, alot more spam is getting past their spam algos, and this is part of the reason why.
_________________
"I stab thee with a rusty spork."

[hornet777]

[Erikalbert]