Predictive Models for Popularity of Solo and Group Singers in Spotify Using Decision Tree

Abstract

Technology has revolutionized our world over the years, and music is one of them. Spotify had become one of the music streaming applications that offers the latest songs in every year. Users can download Spotify from Play Store, App Store, and the official website. Spotify is a music streaming platform that provides a Freemium service, which means it is available for free use and a paid premium. Users can browse the most famous artists or songs easily according to their ratings. The artist, lyric, and some of the audio features have a significant impact on making the song's popularity. The purpose of this research is to predict a soloist or group singers using a decision tree algorithm. We used weighted by information gain method to determine the most crucial feature in processing singers' popularity. Our model will help the solo singer, group singers, and song composers make their songs accessible. The best performance result is the Decision Tree Algorithm with an Accuracy of 89.5% on cross-validation and 92.3% on independent data.