"All of Me": Mining Users' Attributes from their Public Spotify Playlists

Pier Paolo Tricomi,Luca Pajola,Luca Pasa,Mauro Conti

2024-01-26

Abstract:In the age of digital music streaming, playlists on platforms like Spotify have become an integral part of individuals' musical experiences. People create and publicly share their own playlists to express their musical tastes, promote the discovery of their favorite artists, and foster social connections. These publicly accessible playlists transcend the boundaries of mere musical preferences: they serve as sources of rich insights into users' attributes and identities. For example, the musical preferences of elderly individuals may lean more towards Frank Sinatra, while Billie Eilish remains a favored choice among teenagers. These playlists thus become windows into the diverse and evolving facets of one's musical identity.

Cryptography and Security,Machine Learning,Social and Information Networks

What problem does this paper attempt to address?

The paper primarily explores how to mine users' personal attributes by analyzing the playlists they publicly share on the Spotify platform. Specifically, the study aims to address the following questions:

1. **Exploring the relationship between playlists and user attributes**: Researchers want to understand whether there is a connection between users' playlists on Spotify and their personal attributes (such as demographic characteristics, habits, and personality traits).

2. **Identifying playlist features that distinguish different user attribute categories**: Researchers hope to identify which features of playlists, such as songs, artists, and music genres, help distinguish different user attribute categories.

3. **Investigating whether similar users create similar playlists**: The study also aims to examine whether users with similar attributes tend to create similar playlists.

To achieve the above goals, researchers conducted a large-scale online survey and collected data from 739 Spotify users, who provided a total of 10,286 public playlists, involving over 200,000 unique songs and 55,000 artists. Through detailed statistical analysis of this data, the study revealed deep connections between users' playlists and their real-life attributes. For example, individuals with higher openness scores tend to create playlists that include a variety of artists, while female users prefer pop music and K-pop. Based on this, researchers built models that accurately predict user attributes and proposed an application called DeepSet, which performs well in most user attribute prediction tasks.