January 20, 2015
Researchers at Northwestern University, in the United States, have created an automatic method capable of predicting the degree of relevance that a film could have. It combines the measurement of various factors such as reviews, physical sales, and money raised at the box office. It is so accurate that its creators claim that it is more reliable than a movie expert.
The study is based on the collection of films in the United States National Film Registry. There are the most relevant titles of recent years according to a committee of experts. The researchers compared their results with this collection.
To calibrate the algorithm they used data from three different categories: impact, quality and influence. To measure the impact, they used the IMDB website and collected information on user votes for each of the films in the database.
Regarding quality, the data was taken from a website where critics can assign a numerical value to the films. Once with this information, they made a statistical evaluation to determine the importance of the works.
In the results, the films with the greatest influence according to the algorithm were Star Wars, The Wizard of Ozz, Psycho, Casablanca and Gone with the Wind. All these titles are already part of the collection of the National Film Registry. A proof that your system works.
via An algorithm predicts successful movies better than movie critics.
Related articles across the web
[+] Videos de nuestro canal de YouTube