Requião da Cunha B, PhD student - UFRGS Brazilian Federal Police Agent Fast Fragmentation of Networks Using Module-Based Attacks http://dx.doi.org/10.1371/journal.pone.0142824 Gonçalves Sebastia, Proffesor - UFRGS González-Avella JC, IFISC.
Requião da Cunha B,
PhD student - UFRGS
Brazilian Federal Police Agent
Fast Fragmentation of Networks Using Module-Based Attacks
http://dx.doi.org/10.1371/journal.pone.0142824
Gonçalves Sebastia,
Proffesor - UFRGS
González-Avella JC,
IFISC.
Data collection;
• Standardization of qualifiers.
Standardization of qualifiers.
CriminalN = 15.887
EmployN = 2038
FamiliarN = 8495
DoubtsN = 1040
LocalesN = 508
+ ++
++
Snapshot of organized Criminal groups investigated by the Brazilian Federal Police between April and August Of 2013
Snapshot of organized Criminal groups investigated by the Brazilian Federal Police between April and August Of 2013
Module-Based Attacks
Number of Nodes = 9888Number of links = 19744Avg path lenght = 14.43.Clustering Coeficient = 0.428Aveg. degree k= 3.99
Brazilian Criminal Network
Number of Nodes = 9888Number of links = 19744Avg path lenght = 14,43.Clustering Coeficient = 0.428Aveg. degree k= 3.99
Brazilian Criminal Network
Number of modules = 93
Q = 0.95821
Brazilian Criminal Inter-Comunity Network
Number of modules = 93
Inter-modules links = 176
Q = 0.95821
52-(35)-23
linksNodes = 196 Nodes = 77
9
Nodes = 183
(3 links)
52 [35] 2388 [27] 8686 [19] 2433 [13] 8630 [11] 2323 [11] 3121 [11] 6637 [10] 216 [10] 3721 [9] 8928 [8] 636 [8] 2174 [8] 2123 [8] 886 [7] 6721 [7] 9286 [7] 2886 [7] 7252 [6] 3052 [6] 42
28 [5] 4721 [5] 4179 [4] 706 [4] 6632 [4] 5479 [4] 8660 [4] 8652 [4] 3142 [4] 2338 [4] 9236 [3] 4763 [3] 3323 [3] 449 [3] 1524 [3] 8854 [3] 335 [3] 775 [3] 2344 [3] 499 [3] 52
21 [3] 4647 [3] 1415 [3] 219 [3] 438 [2] 317 [2] 3716 [2] 3250 [2] 1972 [2] 2415 [2] 3279 [2] 3385 [2] 1928 [2] 5411 [2] 6333 [2] 7070 [2] 8667 [2] 2482 [2] 2159 [2] 4221 [2] 2
32 [2] 233 [2] 7367 [2] 8815 [2] 4446 [2] 3754 [2] 8633 [2] 6252 [2] 817 [2] 1557 [2] 5491 [2] 659 [1] 3585 [1] 6719 [1] 2046 [1] 6886 [1] 4712 [1] 925 [1] 4915 [1] 333 [1] 2
32 [2] 233 [2] 7367 [2] 8815 [2] 4446 [2] 3754 [2] 8633 [2] 6252 [2] 817 [2] 1557 [2] 5491 [2] 659 [1] 3585 [1] 6719 [1] 2046 [1] 6886 [1] 4712 [1] 925 [1] 4915 [1] 333 [1] 2
Ci CjLink Ci CjLink Ci CjLink Ci CjLink Ci CjLink
Removed Links
Removed Nodes
Fig 2. Comparison between the effect of betweenness-based attack, degree-based attack, longest path attack, random attack, and module-based attack network.
Fig 4. Size of the biggest connected component in terms of the initial size, σ, as function of fraction of removed edges, ρ.
Fig 6. Overall efficiency gain (η) of the MBA method relative to the CBA method as function of modularity, Q, for nodes and edges removal.
The vertical axis is in logarithmic scale and the horizontal axis is linear. The networks attacked are Facebook (FB), Twitter (TW), Google Plus (G+), US power grid (PG), Euro road (ER), Open flights (OF), US airports (UA), Yeast protein (YP), H pylori (HP), and C elegans (CE).
Modularity and the fraction of bridging links.
Data Analisys Tools:
Python,NumpyScipyMatplotlibNetworkxCytoscapeR