Fake online reviews are a plague to sites like Yelp, but never fear, Science is on it: researchers at the University of Illinois at Chicago say they've figured out an algorithm that can detect fake reviews (warning: PDF). Not on a case-by-case basis — that's been researched before — but by looking at groups of fake reviews that are an effort game the system with tons of positive ratings. Hey, that could be pretty useful in getting rid of, say, services that promise to boost your Yelp score.
Anyway, the algorithm looks at things like how often certain accounts review the same places, whether those reviews were posted during the same time period, and how similar the reviews themselves are. Fake reviews, it seems, tend to travel in packs, much like their real-live degenerate, Elite Yelping counterparts.