Document Type

Article

Keywords

Hit Song Science; Bracketology; March Madness; Spotify; YouTube

Identifier Data

https://doi.org/10.48009/4_iis_2021_306-319

Publisher

International Association For Computer Information

Abstract

Each Spring, thousands of middle and high school students enrolled in Spanish classes vote for their favorite songs in the annual Locura De Marzo competition. This alternative March Madness competition gives us an opportunity to build and test models to predict which songs will win which furthers the Hit Song Science literature. Using decision trees and support vector machine (SVM) models we find similarities with the challenge of predicting the popular NCAA Basketball bracket including the importance of seed and the difficulty in predicting a “perfect” bracket

Share

COinS