NYC Gazette

Researchers from New York University have developed an algorithm to identify the traits of Grammy-winning songs. This AI tool analyzes a song's characteristics, such as lyrics and Billboard rankings, to predict winners in categories like Song of the Year, Record of the Year, and Rap Song of the Year for 2021-2023.

“Spotting award-winning art is surely a subjective process and is complicated by the secrecy surrounding voters’ decisions,” says Anasse Bari, a clinical associate professor at NYU’s Courant Institute of Mathematical Science. “However, by taking into account what we know about the songs themselves—from their make-up to their popularity—we can pinpoint those likely to be celebrated.”

The researchers compiled data on nominees from 2004 to 2020 across three categories—Song of the Year, Record of the Year, and Rap Song of the Year—totaling nearly 250 songs. They incorporated variables such as Billboard rankings and Google search volume. Musical characteristics were analyzed using Spotify data, including acousticness, danceability, energy, instrumentalness, and speechiness. The AI also assessed song lyrics through Natural Language Processing algorithms.

The algorithm generated lists of likely winners by identifying top three candidates among nominees for each year studied (2021-2023). It accurately included all nine winning songs across these categories in its top three predictions. For instance, it correctly identified Billie Eilish’s “everything i wanted” (2021 Record of the Year), Silk Sonic’s “Leave the Door Open” (2022 Song of the Year), and Kendrick Lamar’s “The Heart Part 5” (2023 Rap Song of the Year).

Some predictions contradicted betting sites' odds. Bonnie Raitt’s “Just Like That,” which was placed in the top three for 2023 Song of the Year by the model, was considered unlikely to win by gambling platforms. Similarly, H.E.R.’s Grammy-winning “I Can’t Breathe” was viewed as a long shot but was included in the model's top three for 2021 Song of the Year.

Predictive features varied among categories: energy and acousticness were significant for Song of the Year; speechiness and profanity for Record of the Year; vocabulary diversity and happiness score for Rap Song of the Year.

“Our findings highlight the importance of considering multiple factors when predicting music award winners,” says Bari. “More broadly, this work shows potential in using machine learning to gain insights into factors contributing to a song’s success.”

Other authors include Rushabh Musthyala, Abhishek Narayanan, and Anirudh Nistala from NYU’s Department of Computer Science.