Wikipedia talk (fr)
This is the communication network of the French Wikipedia. Nodes represent
users, and an edge from user A to user B denotes that user A wrote a message on
the talk page of user B at a certain timestamp.
Metadata
Statistics
Size  n =  1,420,367

Volume  m =  4,641,928

Unique edge count  m̿ =  2,471,501

Loop count  l =  788,289

Wedge count  s =  623,970,181,277

Claw count  z =  220,098,227,047,176,064

Cross count  x =  6.006 11 × 10^{22}

Triangle count  t =  4,835,877

Square count  q =  11,749,091,667

4Tour count  T_{4} =  2,589,877,992,672

Maximum degree  d_{max} =  1,096,752

Maximum outdegree  d^{+}_{max} =  1,096,720

Maximum indegree  d^{−}_{max} =  60,407

Average degree  d =  6.536 24

Fill  p =  1.225 07 × 10^{−6}

Average edge multiplicity  m̃ =  1.878 18

Size of LCC  N =  1,409,540

Size of LSCC  N_{s} =  56,011

Relative size of LSCC  N^{r}_{s} =  0.039 434 2

Diameter  δ =  11

50Percentile effective diameter  δ_{0.5} =  1.798 42

90Percentile effective diameter  δ_{0.9} =  3.541 98

Median distance  δ_{M} =  2

Mean distance  δ_{m} =  2.592 08

Gini coefficient  G =  0.825 557

Balanced inequality ratio  P =  0.153 816

Outdegree balanced inequality ratio  P_{+} =  0.062 284 7

Indegree balanced inequality ratio  P_{−} =  0.237 932

Relative edge distribution entropy  H_{er} =  0.663 160

Power law exponent  γ =  4.884 49

Tail power law exponent  γ_{t} =  2.601 00

Degree assortativity  ρ =  −0.351 518

Degree assortativity pvalue  p_{ρ} =  0.000 00

In/outdegree correlation  ρ^{±} =  +0.705 697

Clustering coefficient  c =  2.325 05 × 10^{−5}

Directed clustering coefficient  c^{±} =  0.025 045 3

Spectral norm  α =  110,219

Cyclic eigenvalue  π =  55,109.1

Algebraic connectivity  a =  0.024 088 0

Reciprocity  y =  0.139 799

Nonbipartivity  b_{A} =  0.977 206

Normalized nonbipartivity  b_{N} =  0.012 556 7

Spectral bipartite frustration  b_{K} =  0.001 829 73

Controllability  C =  1,351,987

Relative controllability  C_{r} =  0.951 858

Plots
Matrix decompositions plots
Downloads
References
[1]

Jérôme Kunegis.
KONECT – The Koblenz Network Collection.
In Proc. Int. Conf. on World Wide Web Companion, pages
1343–1350, 2013.
[ http ]

[2]

Jun Sun, Jérôme Kunegis, and Steffen Staab.
Predicting user roles in social networks using transfer learning with
feature transformation.
In Proc. ICDM Workshop on Data Min. in Netw., 2016.
