A team of scientists from the Northwestern University of Illinois (USA) has developed an algorithm that, by assessing the quality, influence and impact of a cinematographic work, is able to assign its degree of relevance more reliably than that of an expert film critic – and human . The work has been published in the journal Proceedings of National Academy of Science (PNAS).
Can software rate the greatness of a movie better than a human being when it is so subjective? The algorithm created by these scientists measures its artistic interest, being able “to assign the degree of relevance of a film with greater ease and accuracy than film critics and that, in addition, it can be used in other fields such as the scientific one”, he explains. a Sinc Luís A. Nunes Amaral, co-author of the study.
This tacit sentence is due to the fact that the algorithm has been tested with more than 15,000 films , comparing the results of its system with the National Film Registry of the Library of Congress of the United States (where films as diverse as Bambi , Casablanca, Mary Poppins, Dirty Harry, 2001: A Space Odyssey or The Godfather) .
The algorithm is based on three fundamental factors to measure the relevance of a film: quality (taking into account the opinions of experts from the Metacritic website), impact (with data from the IMDb database) and influence (the network of IMDb citations plus Google PageRank for computer relevance results).
“This method of network analysis can serve other creative endeavors, including scientific research , and may outperform even the best experts in identifying the most significant scientific papers,” Nunes clarifies.
