Well, of course it’s a really good website anyway (it’s helped countless numbers of students to get their degrees, hasn’t it?), but who knew Wikipedia could be used to predict how well a movie will do at the box office too?
Taha Yasseri, a physicist at the Budapest University of Technology and Economics, apparently, who created a mathematical model that processes account data – such as the number of readers for the Wikipedia page of an upcoming movie – and showed that it correlates with box office receipts on the given film’s opening weekend.
Yasseri and two colleagues, Márton Mestyán and János Kertész, built their model using data from 312 movies with Wikipedia pages, out of a total of 535 released in the US in 2010. Overall, the predicted box office takings matched the eventual box office results by 77%.
For bigger movies like Iron Man 2, Toy Story 3 and Inception, the accuracy of the results went up to more than 90%. “We were looking for the fingerprints of popularity of a movie,” said Yasseri, obviously as somebody who spent a whole bunch of time just clicking blue hyperlinks on Wikipedia until the idea struck him.
So whilst some physicists are out there trying to time travel and stuff, other are doing a different kind of research. Unresearch, we’ll call it, because that fits.
Source: The Guardian
Published: Nov 9, 2012 10:39 am