Graph Analysis with Node DifferentialNode Differential PrivacyNode differentially private analysis...
Transcript of Graph Analysis with Node DifferentialNode Differential PrivacyNode differentially private analysis...
-
Graph Analysis with Node Differential PrivacyNode Differential Privacy
SofyaSofya RaskhodnikovaRaskhodnikovaSofyaSofya RaskhodnikovaRaskhodnikovaPenn State University
Joint work with Shiva Shiva KasiviswanathanKasiviswanathan (GE Research),KobbiKobbi NissimNissim (Ben-Gurion U. and Harvard U.),
Ad S i hAd S i h (P S )Adam Smith Adam Smith (Penn State)
1
-
Publishing information about graphs
Many datasets can be represented as graphs• “Friendships” in online social network• “Friendships” in online social network• Financial transactionsE il i ti• Email communication
• Romantic relationships image source http://community.expressor-software.com/blogs/mtarallo/36-extracting-data-facebook-social-graph-expressor-tutorial.html
Privacy is a big issue!
2
American J. Sociology, Bearman, Moody, Stovel
big issue!
-
Private analysis of graph dataGraph G
i )(Government,researchers
Trustedcurator
Users
queries
answers)( researchers,businesses
(or) maliciousmaliciousadversary
T fli i l ili d i• Two conflicting goals: utility and privacy– utility: accurate answers
– privacy: ?
3image source http://www.queticointernetmarketing.com/new-amazing-facebook-photo-mapper/
-
Differential privacy for graph dataGraph G
i )(Government,researchers
Trustedcurator
Users
Aqueries
answers)( researchers,businesses
(or) maliciousmaliciousadversary
• Intuition: neighbors are datasets that differ only in someIntuition: neighbors are datasets that differ only in some information we’d like to hide (e.g., one person’s data)
4image source http://www.queticointernetmarketing.com/new-amazing-facebook-photo-mapper/
-
Two variants of differential privacy for graphs
• Edge differential privacy
GG:
Two graphs are neighbors if they differ in one edge.
• Node differential privacyG:
Two graphs are neighbors if one can be obtained from the other by deleting a node and its adjacent edgesby deleting a node and its adjacent edges.
5
-
Node differentially private analysis of graphsGraph G
i )(Government,researchers
Trustedcurator
Users
Aqueries
answers)( researchers,businesses
(or) malicious
T fli ti l tilit d i
maliciousadversary
• Two conflicting goals: utility and privacy– Impossible to get both in the worst case
• Previously: no node differentially private• Previously: no node differentially private algorithms that are accurate on realistic graphs
6image source http://www.queticointernetmarketing.com/new-amazing-facebook-photo-mapper/
-
Our contributions
• First node differentially private algorithms that are accurate for sparse graphsaccurate for sparse graphs– node differentially private for all graphs
acc rate for a s bclass f h hi h i l d– accurate for a subclass of graphs, which includes • graphs with sublinear (not necessarily constant) degree bound
• graphs where the tail of the degree distribution is not too heavygraphs where the tail of the degree distribution is not too heavy
• dense graphs
• Techniques for node differentially private algorithmsec ques o ode d e e a y p a e a go s
• Methodology for analyzing the accuracy of such algorithms on realistic networksalgorithms on realistic networks
Concurrent work on node privacy [Blocki Blum Datta Sheffet 13]Concurrent work on node privacy [Blocki Blum Datta Sheffet 13]
7
-
Our contributions: algorithms
……
Frequency
…Degrees
8
-
Our contributions: accuracy analysis
(1+o(1))-approximation
9
-
Previous work ondifferentially private computations on graphs
Edge differentially private algorithmsEdge differentially private algorithms• number of triangles, MST cost [Nissim Raskhodnikova Smith 07]• degree distribution [Hay Rastogi Miklau Suciu 09 Hay Li Miklau Jensen 09]• degree distribution [Hay Rastogi Miklau Suciu 09, Hay Li Miklau Jensen 09]• small subgraph counts [Karwa Raskhodnikova Smith Yaroslavtsev 11]• cuts [Blocki Blum Datta Sheffet 12]
Edge private against Bayesian adversary (weaker privacy)ll b h t [ kl ]• small subgraph counts [Rastogi Hay Miklau Suciu 09]
Node zero‐knowledge private (stronger privacy)Node zero knowledge private (stronger privacy)• average degree, distances to nearest connected, Eulerian,
cycle‐free graphs for dense graphs [Gehrke Lui Pass 12]
10
-
Differential privacy basicsGraph G
t ti ti f )(Government,researchers
Trustedcurator
Users
Astatistic f
approximation )( researchers,businesses
(or) maliciousto f(G) maliciousadversary
11
-
Global sensitivity framework [DMNS’06]
12
-
“Projections” on graphs of small degree
Goal: privacy for all graphs
13
-
Method 1: Lipschitz extensions
14
-
1 1'1
s
1
3 3' t
2 2'
1
5 5'
4 4'
15
-
1 1'11/
s
1
3 3' t
2 2'
11/
5 5'
4 4'
16
-
1 1'1
s
1
3 3' t
2 2'
1
5 5'
4 4'
6'6 6'6
17
-
via Smooth Sensitivity framework [NRS’07] via Smooth Sensitivity framework [NRS 07]
18
-
Our results
via Lipschitzextensions
} via generic reduction
19
-
Conclusions
• It is possible to design node differentially private algorithms with good utility on sparse graphswith good utility on sparse graphs– One can first test whether the graph is sparse privately
• Directions for future work– Node‐private synthetic graphs
– What are the right notions of privacy for network data?
20
-
Lipschitz extensions via linear/convex programs
21
-
Our results
……
Frequency
… …Degrees…
22
-
Our results
……
(1+o(1))-approximation(1 o(1)) approximation
23
-
T query fT(G)G A
24S
-
Generic Reduction via TruncationFrequency
… …d Degrees
25
-
Smooth Sensitivity of Truncation
#(nodes of degree above d)
26
-
Releasing Degree Distribution via Generic Reduction
T query fT(G)G A
T q y fT(G)
S
27