Spam Still A Problem. Where's The Backup Plan?

from the no-big-surprise-there dept

The fashionable thing to do these days, if you're an anti-spam company, is come out with a study showing just how little CAN-SPAM did to stop spam. So, here we have yet another study showing that CAN-SPAM has had little to no effect on spamming activities. I'll repeat the same question I asked before the law went into effect: what's the backup plan? In a business, if you are trying to solve a problem, you come up with a plan, and you generally (if you're smart!) have backup plans. You measure how well the first plan does, and make a decision as to whether or not to adjust or go with the backup plan over time. In politics, it seems the methodology is to pass a law, declare the problem solved, pat yourself on the back and forget about it until constituents start complaining again. Here was a case where people should have demanded a clear way to measure the success or failure of the bill along with details on what the backup plan would be. Instead, we've got nothing, but politicians will gladly talk about how they "fought spam!" Meanwhile, John Graham-Cumming, author of the popular POPfile Bayesian spam filter (which I use as one of many spam fighting techniques) has figured out that there are certain key words that a spammer can include in spam to defeat a Bayesian filter every time. The good news, though, is that those keywords are different for each person. This isn't a huge surprise of course. Legitimate (usually work or friend related) emails make it through a filter because of a few major words that tend to prove legitimacy. But, those words will differ based on your job, friends and hobbies. Still, the article includes the words that work for Graham-Cumming, and I wonder how many spammers will misread the article and assume those keywords will work for everyone. Expect to see those words showing up a lot more in your spam in the near future.