Math in the News
IMDb Buzz Used in Oscar Predictions
A mathematical model incorporating such factors as the intensity and positivity of the chatter about a film in the Internet Movie Database (IMDb) comments section has proven successful at predicting Academy Award nominees.
Developed by, among others, Peter Gloor of the Center for Collective Intelligence and MIT's Sloan School of Management, the model predicted the 2012 nominees for Best Picture and Best Director, along with about 70 percent of the acting nominees.
The difficulty with actors, Gloor explains, is that buzz about them isn't always relevant to their favor with the Academy: "Directors, they are not so active in their love life—they are speculated about if it's an Oscar, not if it's a break-up with a boyfriend or girlfirend."
Gloor's Oscar research is part of a larger body of work on "swarm creativity," or the "wisdom of crowds."
Read the Vanity Fair story or the full text of the academic paper.