This project dives deep into the relation between the various latent attributes of the music answering the psychological factors behind peoples' different music preferences. Music is heard by people daily in many parts of the world, and affects people in various ways from emotion regulation to cognitive development, along with providing a means for self-expression. Music training has been shown to help improve intellectual development and ability.
There was a need at start of this project to find a dataset giving out various audio features of particular track. After the lot of research over Internet to find the reliable source, it was found that Spotify which is a music, podcast, and video streaming service that was officially launched on 7 October 2008. It was developed by Spotify AB in Stockholm, Sweden. It provides DRM–protected content from record labels and media companies. Spotify is a freemium service; basic features are free with advertisements or limitations, while additional features, such as improved streaming quality and music downloads, are offered via paid subscriptions.
Spotify Web Api provides various https://developer.spotify.com/web-api/get-audio-features/ of each track which is provided in the hyper link. Oberserving the diversity of audio features provided by the Spotify a huge potential was seen to analyze further to find the complex relation between them.
The other challenge was to accumulate a large diverse dataset by bulk calling the Spotify API to provide the audio features for this data mining project. The R package Spotifr which is spotifyr is a quick and easy wrapper for pulling track audio features from Spotify’s Web API in bulk. By automatically batching API requests, it allows to enter an artist’s name and retrieve their entire discography in seconds, along with Spotify’s audio features and track/album popularity metrics.