RDF for mail filtering: FOAF whitelists
What you do is generate a checksum based on your mailbox ID. Then you distribute that checksum, and people create “whitelists” (but let's call 'em goodlists, instead, okay, and get that color thing out of there, time for a little accuracy in our semantics) with your checksums - NOT your email address, that can't be guessed by looking at the checksum. And you add a list of your own addresses, see, everyone shares “good” email addresses in a big database and that way mail that comes from the peoples, the good peoples, can be flagged higher, and mail that comes from spam folks gets flagged lower. There you have it, a real application of the Semantic Web, although it doesn't fully exist yet, but yep.