genius_expertise — Genius users, annotations and artists (2019-2020)

Description

Users, songs and artists of the lyrics annotation website Genius (genius.com), crawled between September 2019 and January 2020 by D. Lim and A. R. Benson to study the expertise and dynamics of crowdsourced knowledge curation. On Genius, users write annotations explaining passages of song lyrics; annotations are edited collaboratively and reviewed by moderators, and verified annotations are written by the artists themselves. Users can follow each other. There are three networks:

follows: a directed network of users, where an edge $i \to j$ means that user $i$ follows user $j$ (users who appear only in the user information file are included as isolated nodes; 1,219 users follow themselves, as given).
annotations: a bipartite multigraph of users and songs, where each edge is an edit of an annotation of the song's lyrics by the user. Reviewed annotations come with their edit histories, where each edit is an edge; unreviewed annotations have only their author, and verified annotations only the artist account that wrote them, each of which is a single edge with the time of the annotation.
artist_songs: a bipartite network of artists (as given by their Genius artist pages) and the songs they contributed to (e.g. as performers, writers or producers).

User names and the names of songs and artists are the identifiers of their pages on the site (e.g. https://genius.com/<name>); artist and user accounts are distinct, so that artists in artist_songs are not identified with users. The user information file lists 2,513 users more than once, and the artist information file 2,543 artists, in part with different values; the first entry is used (for the artists, with the songs of all their entries). The times are given without time zone, and are interpreted as UTC. Left out are the texts of the annotations and their edits, the lyrics of the songs, the contribution shares of the users to each reviewed annotation computed by Genius, and the users who accepted the annotations.

The networks have the following properties:

(Node) name: the name of the user, song or artist page.
(Node) is_song (annotations, artist_songs): whether the node is a song.
(Node) iq (users, in follows and annotations): the user's IQ (Genius reputation points); -1 for users without information.
(Node) roles (users, in follows and annotations): the roles of the user (e.g. "Editor", "Contributor").
(Node) annotations_num, suggestions_num, transcriptions_num, questions_num, answers_num, pyongs_num (users, in follows and annotations): the numbers of annotations, suggestions, transcriptions, questions, answers and pyongs (shares) of the user; -1 for users without information.
(Node) followers_num, following_num (users, in follows and annotations): the numbers of followers and of followed users; -1 for users without information.
(Node) title, primary_artist, release_date, views, pyongs (songs, in annotations and artist_songs): the title, primary artist, release date (as given, e.g. "July 31, 2012"), views and pyongs of the song; empty or -1 for songs without information.
(Node) contributors (songs, in annotations and artist_songs): the number of contributors; -1 for songs without information.
(Node) has_bio (songs, in annotations and artist_songs): whether the song has a description.
(Node) tags (songs, in annotations and artist_songs): genre tags; empty for songs without information.
(Node) iq, followers, roles (artists, in artist_songs): the IQ, followers and roles of the artist.
(Edge) time (annotations): the time of the edit (or of the annotation), in Unix time.
(Edge) annotation (annotations): the numeric identifier of the annotation.
(Edge) type (annotations): the type of the annotation ("reviewed", "unreviewed" or "verified").
(Edge) votes (annotations): the votes of the annotation.

Tags

Social Online Collaboration Multigraph Timestamps Unweighted Metadata

Added

2026-10-27

Citation

D. Lim and A. R. Benson, "Expertise and Dynamics within Crowdsourced Musical Knowledge Curation: A Case Study of the Genius Platform." Proceedings of the International AAAI Conference on Web and Social Media (ICWSM) 15, 373-384 (2021). 10.1609/icwsm.v15i1.18068 Sci-Hub

Upstream URL OK

https://www.cs.cornell.edu/~arb/data/genius-expertise/
https://github.com/cptq/genius-expertise

Networks

Tip: click on a column header to sort the list, or hover over it for a description.

Column legend

$\left<k\right>$: Mean degree
$\sigma_k$: Standard deviation of the degrees
$\lambda_h$: Largest eigenvalue of the Hashimoto (non-backtracking) matrix
$\tau$: Random walk mixing time
$r$: Degree assortativity
$c$: Global clustering coefficient
$\oslash$: (Pseudo-) diameter
$S$: Relative size of the largest component
NPs: Node properties
EPs: Edge properties

Name	Nodes	Edges	$\left<k\right>$	$\sigma_k$	$\lambda_h$	$\tau$	$r$	$c$	$\oslash$	$S$	Kind	Mode	NPs	EPs	gt	GraphML	GML	csv
follows	780,903	1,778,595	2.28	58.69	253.86	2225.42	-0.10	0.01	16	1.00	Directed	Unipartite	name annotations_num suggestions_num transcriptions_num questions_num answers_num pyongs_num iq followers_num following_num roles		20.2 MiB	29.8 MiB	29.4 MiB	25.3 MiB
annotations	119,310	1,263,717	21.18	140.93	67.36	159.43	-0.09	0.00	16	0.97	Undirected	Bipartite	name is_song annotations_num suggestions_num transcriptions_num questions_num answers_num pyongs_num iq followers_num following_num roles title primary_artist release_date views pyongs contributors has_bio tags	time annotation type votes	11.1 MiB	16.2 MiB	15.4 MiB	15.3 MiB
artist_songs	1,994,956	3,184,366	3.19	18.74	48.76	5228.08	-0.15	0.00	28	0.88	Undirected	Bipartite	name iq followers roles is_song title primary_artist release_date views pyongs contributors has_bio tags		24.9 MiB	37.8 MiB	36.9 MiB	37.1 MiB