Filmtipset comments
This is the bipartite network of users commenting on movies, from the Swedish
movie rating website Filmtipset.se. The nodes of the network are the users and
the movie. Each link connects a user with a movie, and denote a single comment
by that users on that movie. Multiple link between the same user–movie pair
are allowed. Each link is annotated with the timestamp of the comment.
Metadata
Statistics
Size  n =  75,360

Left size  n_{1} =  29,530

Right size  n_{2} =  45,830

Volume  m =  1,266,753

Unique edge count  m̿ =  1,204,831

Wedge count  s =  493,367,139

Claw count  z =  204,725,098,142

Cross count  x =  136,282,597,862,835

Square count  q =  2,724,858,975

4Tour count  T_{4} =  23,774,988,546

Maximum degree  d_{max} =  5,659

Maximum left degree  d_{1max} =  5,659

Maximum right degree  d_{2max} =  2,252

Average degree  d =  33.618 7

Average left degree  d_{1} =  42.897 2

Average right degree  d_{2} =  27.640 3

Fill  p =  0.000 890 252

Average edge multiplicity  m̃ =  1.051 39

Size of LCC  N =  75,194

Diameter  δ =  9

50Percentile effective diameter  δ_{0.5} =  3.297 78

90Percentile effective diameter  δ_{0.9} =  4.305 01

Mean distance  δ_{m} =  3.749 21

Gini coefficient  G =  0.803 195

Balanced inequality ratio  P =  0.171 261

Left balanced inequality ratio  P_{1} =  0.169 335

Right balanced inequality ratio  P_{2} =  0.144 105

Relative edge distribution entropy  H_{er} =  0.851 035

Power law exponent  γ =  1.578 01

Tail power law exponent  γ_{t} =  1.581 00

Degree assortativity  ρ =  −0.186 575

Degree assortativity pvalue  p_{ρ} =  0.000 00

Spectral norm  α =  358.889

Algebraic connectivity  a =  0.196 326

Spectral separation  λ_{1}[A] / λ_{2}[A] =  2.351 83

Controllability  C =  40,256

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]

Alan Said, Ernesto W. De Luca, and Sahin Albayrak.
How social relationships affect user similarities.
In Proc. IUI Workshop on Soc. Recomm. Syst., 2010.
