One of the more consistent pains across every project I’ve been on is data.  When you get down to it, every little app and giant enterprise integration is fueled by  bits of characters and numbers; they’re split into pieces, handled by lots of people, and are somehow reconstructed to provide context.

While a user can innately tell if this context “makes sense,” it takes a very mature set of rules and scenarios to glean if having your appointment at 3 AM makes sense.  How can we get to the next level?  Do we use bayesian filtering and classification to fake learning, or can we start making predictions about what the user “wants.”
I’m not really sure, but there’s a lot of money out there waiting for someone that comes up with a succinct way to handle scheduling a global conference for a time when everyone’s available and awake.

Every year, English teachers from across the country submit their collections of actual analogies and metaphors found in high school essays. These excerpts are published each year to the amusement of teachers across the country.

Here are last year’s winners:

http://www.munsinger.com/malaprop.html