Professional Documents
Culture Documents
ARS-CH4-Local Community Identification
ARS-CH4-Local Community Identification
ARS-CH4-Local Community Identification
Rushed Kanawati
LIPN, CNRS UMR 7030; USPC
http://lipn.fr/∼kanawati
rushed.kanawati@lipn.univ-paris13.fr
1 / 32
I NTRODUCTION Community detection
P LAN
1 I NTRODUCTION
2 Community detection
Local community detection
2 / 32
I NTRODUCTION Community detection
C OMPLEX NETWORK
Definition
Graphs modeling (direct/indirect) interactions among actors.
I Community structure
3 / 32
I NTRODUCTION Community detection
C OMMUNITY ?
Definitions
I A dense subgraph loosely coupled to other modules in the
network
I A community is a set of nodes seen as one by nodes outside the
community
I A subgraph where almost all nodes are linked to other nodes in
the community.
I ...
4 / 32
I NTRODUCTION Community detection
5 / 32
I NTRODUCTION Community detection
L OCAL COMMUNITY
6 / 32
I NTRODUCTION Community detection
L OCAL COMMUNITY
7 / 32
I NTRODUCTION Community detection
L OCAL COMMUNITY
8 / 32
I NTRODUCTION Community detection
9 / 32
I NTRODUCTION Community detection
Blank nodes enhance Bin and Din without affecting Bout and Dout
Bank nodes will be added if M or R modularities are used
Low precision computed communities
Proposed solution: Ensemble approaches
10 / 32
I NTRODUCTION Community detection
11 / 32
I NTRODUCTION Community detection
Principle
Let Qi (s) be the local modularity value induced by node s ∈ S
Qi (s)−min Qi (u)
u∈S
if max Qi (u) 6= min Qi (u)
] max Qi (u)−min Qi (u)
Qi (s) = u∈S u∈S u∈S u∈S
1
otherwise
k
1 P ]
Qcom (s) = k Qi (s)
i=1
12 / 32
I NTRODUCTION Community detection
Principle
I Rank S in function of each local modularity Qi
I Select the winner after applying ensemble ranking approach
I What stopping criteria to apply ?
Stopping criteria
I Strict policy : All modularities should be enhanced
I Majority policy : Majority of local modularities are enhanced
I Least gain policy : At least one local modularity is enhanced.
13 / 32
I NTRODUCTION Community detection
E NSEMBLE RANKING
Problem
I Let S be a set of elements to rank by n rankers
I Let σi be the rank provided by ranker i
I Goal: Compute a consensus rank of S.
14 / 32
I NTRODUCTION Community detection
E NSEMBLE RANKING
Problem
I Let S be a set of elements to rank by n rankers
I Let σi be the rank provided by ranker i
I Goal: Compute a consensus rank of S.
15 / 32
I NTRODUCTION Community detection
Borda
I Borda’s score of i ∈ σk :
Bσk (i) = {count(j)|σk (j) < σk (i) ; j ∈ σk }.
I Rank elements in function of
B(i) = kt=1 wt × Bσt (i).
P
Jean-Charles de Borda [1733-1799]
16 / 32
I NTRODUCTION Community detection
Condorcet
I The winner is the candidate who defeats
every other candidate in pairwise
majority-rule election
I The winner may not exists
Marquis de Condorcet [1743-1794]
17 / 32
I NTRODUCTION Community detection
Condorcet
I The winner is the candidate who defeats
every other candidate in pairwise
majority-rule election
I The winner may not exists
Marquis de Condorcet [1743-1794]
Condorcet 6= Borda
I Votes : 6 × A B C, 4 × B C A
I Borda winner : B
I Condorcet winner : A
18 / 32
I NTRODUCTION Community detection
Arrow’s Theorem
No vote rule can have the following desired
proprieties :
I Every result must be achievable somehow.
I Monotonicity.
I Independence of irrelevant attributes.
Kenneth Arrow, 1921-
I Non-dictatorship.
19 / 32
I NTRODUCTION Community detection
I NP-Hard problem
20 / 32
I NTRODUCTION Community detection
Principle
I Let CQ i
vq be the the local community of vq applying Qi .
21 / 32
I NTRODUCTION Community detection
from A. Topchy et. al. Clustering Ensembles: Models of Consensus and Weak Partitions. PAMI, 2005
22 / 32
I NTRODUCTION Community detection
23 / 32
I NTRODUCTION Community detection
CSPA : E XEMPLE
from Seifi, M. Cœurs stables de communautés dans les graphes de terrain. Thèse de l’université Paris 6, 2012
24 / 32
I NTRODUCTION Community detection
I Complexity : O(nkr)
25 / 32
I NTRODUCTION Community detection
I Complexity : O(nk2 r2 )
26 / 32
I NTRODUCTION Community detection
E XPERIMENTS
Protocol ([Bag08])
1 Apply the different algorithms on nodes in networks for which a
ground-truth community partition is known.
27 / 32
I NTRODUCTION Community detection
D ATASETS
28 / 32
I NTRODUCTION Community detection
29 / 32
I NTRODUCTION Community detection
30 / 32
I NTRODUCTION Community detection
B IBLIOGRAPHY I
J. P. Bagrow, Evaluating local community methods in networks, J. Stat. Mech. 2008 (2008), no. 5, P05001.
Aaron Clauset, Finding local community structure in networks, Physical Review E (2005).
Jiyang Chen, Osmar R. Zaı̈ane, and Randy Goebel, Local community identification in social networks, ASONAM, 2009,
pp. 237–242.
Feng Luo, James Zijun Wang, and Eric Promislow, Exploring local community structures in large networks, Web Intelligence
and Agent Systems 6 (2008), no. 4, 387–400.
Marina Meila, Comparing clusterings by the variation of information, COLT (Bernhard Schölkopf and Manfred K. Warmuth,
eds.), Lecture Notes in Computer Science, vol. 2777, Springer, 2003, pp. 173–187.
31 / 32
I NTRODUCTION Community detection
B IBLIOGRAPHY II
Jaewon Yang and Jure Leskovec, Defining and evaluating network communities based on ground-truth, ICDM
(Mohammed Javeed Zaki, Arno Siebes, Jeffrey Xu Yu, Bart Goethals, Geoffrey I. Webb, and Xindong Wu, eds.), IEEE
Computer Society, 2012, pp. 745–754.
32 / 32