← Newer   Older →   ↑ Back to List

Spam: Gods, such wastefulness

POPFile is now trained and works magnificently. It’s now 82.93% accurate, up from about 75% a few weeks ago, and that’s good—if it’s climbing, it means it’s making fewer mistakes, which is certainly true.

(For those of you who have no idea what this post is about, you might want to read up on Bayesian filtering”:http://www.paulgraham.com/spam.html from “Paul Graham Then come back, and try using POPFile yourself. It’s worth it, even if you only get five or six emails a day, because if even one of those is junk you’re wasting time.)

I receive 75-100 emails a day, generally. Sometimes as few as 50. (Phew!) I only have statistics for about the past month, however, because that’s when I lost my last POPFile corpus in the scuffle with my computer, and so I don’t have that many emails to go on.

Additionally, my mailing lists have what’s called a magnet, which means that mail I receive on them goes straight into the ‘mail’ bucket without reading through its contents to see if it’s junk or not. (This works really well, as I don’t receive any spam on the lists and this reduces my false positives. Substantially.) So let’s take away maybe 40 emails a day, out of my 75-email average.

I now have about 35 emails a day to go on. That figure feels about right, since I have classified about 725 emails. Let me see — it’s 717 to be precise. Of those emails, 527 have turned out to be junk and 190 have been mail. 75% of my email, roughly, is junk. 75 PERCENT! That’s mind-numbing! And I get a lot of email. For someone who doesn’t get that much, like my mother, say, or my father, they’re probably looking at well over 90% of their email as spam. They currently have about 2000 spam emails accumulated on my dad’s computer downstairs. (They manually filter them. I’ll never talk them into Bayesian filtering.)

Breathtaking, absolutely breathtaking! The waste! Think about all the bandwidth required to convey that many spam messages around the world, to waste my time downloading them especially when I’m already running a VPN(Virtual Private Network) from Portland to Evanston so I can access NU-specific resources (which means that in fact, I am wasting Northwestern’s bandwidth and my own), my just to clot my inbox.

Makes you want to scream. And then, make it so illegal that you can practically put a spammer to death. I’m opposed to the death penalty, but here’s a time when I might make an exception.

commenting closed for this article