Beats and Bytes: Decoding Song’s Success Patterns Using Machine Learning

With the progress of technologies and advent of the likes of Big Data, storing burgeoning volumes of data has become an easy and minuscule task. As the basic paradigm of computer science goes, this data remains latent until and unless processed. Spotify, the online streaming service, has become a household name globally. However, most of its revenue is expended as licensing fee to the big labels in exchange for streaming their music on Spotify’s platform. This study attempts to convert data into information by exploiting the patterns in audio features. It tries to determine whether a song will make it to the Top-50 chart by leveraging past data, thus predicting whether certain audio characteristics of a song are highly correlated to its eventual success in the market. The proposed work would enable artists and labels to decide what makes a "Hit" song more objectively, thus blurring the line between creativity and technology.