So the week of graduation, a friend casually asks me, "Hey Herscher, why weren't you at the nerdhouse graduation party the other night?" Huh? Party? Was I not invited to a party? Wait, that can't be it.. I'm on the nerdhouse-party mailing list! What the heck is going on?
Sure enough, it turns out Gmail filtered the invitation (a relatively normal-looking email sent by a human) as spam. I wrote a paper on this topic -- "false negatives" in spam filtering processes -- for my EECS 547 E-Commerce course a few semesters back (and I'd post it if it weren't on a hard drive buried in shipping boxes on their way to Seattle right about now).
Anyhow, it got me thinking.. what else has Gmail mistakenly filtered as spam? I have no way of finding out. My Spam folder has 5,372 items from the last 30 days. That's over 107 pages of message headings.
So, I decided to do a quick search through my spam folder of messages sent from addresses ending with @*umich.edu. Sure enough, there were more messages mistakenly filtered as spam.
There's no silver bullet (yet) when it comes to fighting spam. But I'm beginning to think that Gmail might be a bit overzealous with their filtering methods. There's also no way to change the threshold (at least, no way that I can seem to find). And, there's no way to sort my spam by % chance that it is in fact spam, or any other way of intelligently sorting through mail filtered as spam.
The fact that only 1 of about 500 spam messages slips through the filters into my Inbox every couple of days makes me rather suspicious of what other legitimate email I'm missing.
Anyhow, don't take this as "anti-Google" or "anti-Gmail". I use Gmail because it's the best web-based email system I've seen so far (the fact that it's free is nice too). I hope they keep improving it (and Blogger too, damnit! port over the gmail spellchecker already!).