Wikipedia talk (it)
This is the communication network of the Italian 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 =  863,846

Volume  m =  3,067,680

Unique edge count  m̿ =  1,661,453

Loop count  l =  233,216

Wedge count  s =  111,830,205,348

Claw count  z =  11,983,312,889,964,536

Cross count  x =  1.065 98 × 10^{21}

Triangle count  t =  3,355,399

Square count  q =  6,154,887,250

4Tour count  T_{4} =  496,562,923,582

Maximum degree  d_{max} =  388,889

Maximum outdegree  d^{+}_{max} =  388,882

Maximum indegree  d^{−}_{max} =  15,811

Average degree  d =  7.102 38

Fill  p =  2.226 46 × 10^{−6}

Average edge multiplicity  m̃ =  1.846 38

Size of LCC  N =  862,214

Size of LSCC  N_{s} =  36,356

Relative size of LSCC  N^{r}_{s} =  0.042 086 2

Diameter  δ =  7

50Percentile effective diameter  δ_{0.5} =  2.554 95

90Percentile effective diameter  δ_{0.9} =  3.670 68

Median distance  δ_{M} =  3

Mean distance  δ_{m} =  3.049 93

Gini coefficient  G =  0.832 891

Balanced inequality ratio  P =  0.149 587

Outdegree balanced inequality ratio  P_{+} =  0.064 084 9

Indegree balanced inequality ratio  P_{−} =  0.232 251

Relative edge distribution entropy  H_{er} =  0.681 910

Power law exponent  γ =  4.229 38

Tail power law exponent  γ_{t} =  1.791 00

Degree assortativity  ρ =  −0.301 420

Degree assortativity pvalue  p_{ρ} =  0.000 00

In/outdegree correlation  ρ^{±} =  +0.694 897

Clustering coefficient  c =  9.001 32 × 10^{−5}

Directed clustering coefficient  c^{±} =  0.030 690 5

Spectral norm  α =  9,318.44

Operator 2norm  ν =  4,750.84

Cyclic eigenvalue  π =  4,599.85

Algebraic connectivity  a =  0.102 767

Spectral separation  λ_{1}[A] / λ_{2}[A] =  1.103 69

Reciprocity  y =  0.174 273

Nonbipartivity  b_{A} =  0.712 796

Normalized nonbipartivity  b_{N} =  0.029 882 2

Spectral bipartite frustration  b_{K} =  0.004 722 98

Controllability  C =  831,096

Relative controllability  C_{r} =  0.962 088

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.
