Every year, you hear a lot of complaints about the films that did and didn't receive a lot of love in the Oscar nominations. While this year is no different, let's not get too worked up over the whole thing. Apparently, a computer might be just as successful at naming the best movies.
Computer analysis may be just as good or even better at identifying significant films as movie experts, according to a new study from Northwestern University's McCormick School of Engineering. Researchers used data from IMDb to create a system that tracked a film's influence over history to try and predict which movies the U.S. Library of Congress will add to the National Film Registry.
The data included the number of "Connections" listed for each film on IMDb, the movies, TV shows and other works that reference it. The study, published in the journal Proceedings of the National Academy of Sciences of the United States of America Monday, looked at 15,425 films on IMDb.
This algorithm had a 91 percent accuracy in predicting which films would be added to the registry based on the number of references to them more than 25 years after their release. Critics' reviews, Oscar wins and box office earnings didn't predict the films added to the National Film Registry nearly as well, according to the study.
Every year, the Librarian of Congress chooses 25 movies that are "culturally, historically or aesthetically" significant to be added to the National Film Registry. With that criteria in mind, it makes sense that the films with more mentions in pop culture throughout history would be more likely to be added to this registry.
However, the algorithm isn't perfect. According to the system used in the study, the 1971 film Willy Wonka and the Chocolate Factory should have already been added to the National Film Registry. The film has 52 log-lag citations to it on IMDb, the 37th most in the study. I think we can all agree that it's pretty criminal that this movie hasn't been added to this prestigious collection yet. Wrong sir, wrong!
"Experts have biases that can affect how they evaluate things," the study's co-author Luis A.N. Amaral told Reuters. "Automated, objective methods don't suffer from that. It may hurt our pride, but they can perform as well as or better than experts."
Of course, Willy Wonka and the Chocolate Factory could always be added to the National Film Registry next year, as could Dumbo, Spartacus or The Shining, other films the algorithm predicted would make the list. There's no earthly way of knowing.
