In this project, I will be performing an analysis of the taste of the listeners on platforms like Spotify or Apple music or such and would try to mark all important insights from harmonic features of today’s music like timbres, usage of major, minor or pentatonic scales used in the production of music to even simpler aspects of music like whether the songs have some catchy beats or melodies or like how much proportion of listeners would like to dance or sing along the song. All of the following features would be considered and this will help to even address what proportion of listeners like to hear a particular artist or a song. This all will be performed using the link prediction with input as the graph content and contents and features of the song as well to calculate the highest accuracy scores which will further help me in the analysis proposed.
BillBoard - Billboard is an American music and entertainment magazine published weekly by the Billboard-Hollywood Reporter Media Group, a division of MRC Media & Info. The magazine provides music charts, news, video, opinion, reviews, events, and style related to the music industry.
Spotify - Spotify is a Swedish audio streaming and media services provider founded on 23 April 2006 by Daniel Ek and Martin Lorentzon. It is the world's largest music streaming service provider, with over 381 million monthly active users, including 172 million paying subscribers, as of September 2021.
CS224W Final Report: Influence Networks in Popular Music Marco Alban, Vivek Choksi, Stephanie Tsai, Stanford CS Department. Link
The Dataset used is a Demographic-based Top Artist In A Country Dataset where the dataset contains about 53 Countries in a form of dictonary keys with values as top 10 music artist in the given region for a particular year given. The Datset was extracted from one the research work done with the same base paper as mine and I have also included it in the Literature survey. The data were all collected from Last.fm API.