FilmTrust ratings
This is bipartite rating network of the FilmTrust project. We removed three
instances of duplicate user–film pairs; making the network simple rather than
having multiple edges.
Metadata
Statistics
Size  n =  3,579

Left size  n_{1} =  1,508

Right size  n_{2} =  2,071

Volume  m =  35,494

Wedge count  s =  9,820,232

Claw count  z =  2,067,481,321

Cross count  x =  376,620,963,773

Square count  q =  99,312,706

4Tour count  T_{4} =  833,882,216

Maximum degree  d_{max} =  1,044

Maximum left degree  d_{1max} =  244

Maximum right degree  d_{2max} =  1,044

Average degree  d =  19.834 6

Average left degree  d_{1} =  23.537 1

Average right degree  d_{2} =  17.138 6

Fill  p =  0.011 365 1

Size of LCC  N =  3,574

Diameter  δ =  7

50Percentile effective diameter  δ_{0.5} =  2.739 74

90Percentile effective diameter  δ_{0.9} =  3.839 54

Median distance  δ_{M} =  3

Mean distance  δ_{m} =  3.255 39

Gini coefficient  G =  0.716 391

Balanced inequality ratio  P =  0.232 969

Left balanced inequality ratio  P_{1} =  0.312 194

Right balanced inequality ratio  P_{2} =  0.116 076

Relative edge distribution entropy  H_{er} =  0.811 239

Power law exponent  γ =  1.605 27

Tail power law exponent  γ_{t} =  1.601 00

Tail power law exponent with p  γ_{3} =  1.601 00

pvalue  p =  0.000 00

Left tail power law exponent with p  γ_{3,1} =  3.181 00

Left pvalue  p_{1} =  0.246 000

Right tail power law exponent with p  γ_{3,2} =  2.061 00

Right pvalue  p_{2} =  0.000 00

Degree assortativity  ρ =  −0.450 722

Degree assortativity pvalue  p_{ρ} =  0.000 00

Spectral norm  α =  45.347 7

Algebraic connectivity  a =  0.007 217 86

Spectral separation  λ_{1}[A] / λ_{2}[A] =  1.343 74

Negativity  ζ =  0.439 736

Algebraic conflict  ξ =  0.267 701

Spectral signed frustration  φ =  0.003 369 74

Controllability  C =  2,478

Relative controllability  C_{r} =  0.692 372

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]

Guobing Guo, Jia Zhang, and Neil YorkeSmith.
A novel Bayesian similarity measure for recommender systems.
In Proc. Int. Joint Conf. on Artif. Intell., pages 2619–2625,
2013.
