This is the social network of YouTube users and their friendship connections.


Internal nameyoutube-u-growth
Data source
AvailabilityDataset is available for download
Consistency checkDataset passed all tests
Online social network
Node meaningUser
Edge meaningFriendship
Network formatUnipartite, undirected
Edge typeUnweighted, no multiple edges
Temporal data Edges are annotated with timestamps
LoopsDoes not contain loops


Size n =3,223,589
Volume m =9,375,374
Wedge count s =26,535,309,102
Claw count z =409,390,882,186,801
Cross count x =6,774,196,341,151,129,600
Triangle count t =12,226,580
Square count q =9,890,851,109
4-Tour count T4 =185,286,796,028
Maximum degree dmax =91,751
Average degree d =5.816 73
Fill p =1.804 43 × 10−6
Size of LCC N =3,216,075
Diameter δ =31
50-Percentile effective diameter δ0.5 =4.652 66
90-Percentile effective diameter δ0.9 =6.643 30
Mean distance δm =5.291 29
Gini coefficient G =0.728 959
Relative edge distribution entropy Her =0.859 534
Power law exponent γ =2.337 95
Tail power law exponent γt =2.211 00
Degree assortativity ρ =−0.063 163 4
Degree assortativity p-value pρ =0.000 00
Clustering coefficient c =0.001 382 30
Spectral norm α =464.866
Non-bipartivity bA =0.082 089 1


Degree distribution

Cumulative degree distribution

Lorenz curve

Spectral distribution of the adjacency matrix

Spectral distribution of the normalized adjacency matrix

Spectral distribution of the Laplacian

Spectral graph drawing based on the adjacency matrix

Spectral graph drawing based on the normalized adjacency matrix

Hop distribution

Temporal distribution

Diameter/density evolution



[1] Jérôme Kunegis. KONECT – The Koblenz Network Collection. In Proc. Int. Conf. on World Wide Web Companion, pages 1343–1350, 2013. [ http ]
[2] Alan Mislove. Online Social Networks: Measurement, Analysis, and Applications to Distributed Information Systems. PhD thesis, Rice Univ., 2009.