Last.Fm UK User Graph Dataset: A Social Network and Music Listening Behavior Dataset

The Last.Fm UK User Graph Dataset is a comprehensive collection of social network and music listening behavior data obtained from the Last.Fm platform. The dataset includes user connections, weekly artist listening counts, and artist tags weighted by the number of users who assigned them. The dataset is ideal for studying social network dynamics, music preferences, and their interplay. 

Dataset Description:

Data Source:

Last.Fm APIs were used to obtain a sample of the UK user graph, exploring the network with a breadth-first approach up to the fifth degree of separation from the seeds.

Social Network Data:

The social graph (G) consists of 75,969 nodes representing users and 389,639 edges representing their friendships on the Last.Fm platform.

Music Listening Behavior Data:

For each user, the dataset includes the number of single listenings of a given artist for each week in the time window from Jan-10 to Dec-11.

Artist Tags:

Tags assigned to artists are included, weighted by the number of users who assigned the tag. Tags were split into single words, filtered for musical genres, and assigned to artists based on specific criteria.

Musical Genre Assignment:

Musical genres were assigned to artists based on the survived tag with the greatest counter, with a relative rate ≥ 0.5.

Files Description:

ArtistsMap: semicolon-separated file that maps artists names to unique ids

ArtistTag: semicolon-separated file with tags related to each artist and their weight (i.e. number of users assigning that tag to that artist)

network: semicolon-separated file containing the social graph edge list

UsersData_anonymized: semicolon-separated file containing user-related attributes