the hardness of being private
play

The Hardness of Being Private Anil Ada, Arkadev Chattopadhyay, - PowerPoint PPT Presentation

The Hardness of Being Private Anil Ada, Arkadev Chattopadhyay, Stephen Cook, Lila Fontes, Michal Kouck y, Toniann Pitassi IEEE Conference on Computational Complexity 2012 Porto, Portugal Lila Fontes (University of Toronto) 0 / 12


  1. The Hardness of Being Private Anil Ada, Arkadev Chattopadhyay, Stephen Cook, Lila Fontes, Michal Kouck´ y, Toniann Pitassi IEEE Conference on Computational Complexity 2012 Porto, Portugal Lila Fontes (University of Toronto) 0 / 12

  2. Communication complexity Two-player model each player has a private input (Alice has x ∈ X , Bob has y ∈ Y ) players communicate over a channel players follow a protocol to compute f : X × Y → Z the last message sent is the value f ( x , y ) = z Lila Fontes (University of Toronto) 1 / 12

  3. Communication complexity Two-player model each player has a private input (Alice has x ∈ X , Bob has y ∈ Y ) players communicate over a channel players follow a protocol to compute f : X × Y → Z the last message sent is the value f ( x , y ) = z The communication cost of a protocol is the worst-case length of the full transcript. Lila Fontes (University of Toronto) 1 / 12

  4. Communication complexity The model Matrix M f has entries M f [ x , y ] = f ( x , y ). Lila Fontes (University of Toronto) 2 / 12

  5. Communication complexity The model Matrix M f has entries M f [ x , y ] = f ( x , y ). A submatrix is monochromatic if f is constant on inputs in the submatrix. Lila Fontes (University of Toronto) 2 / 12

  6. Communication complexity The model Matrix M f has entries M f [ x , y ] = f ( x , y ). A submatrix is monochromatic if f is constant on inputs in the submatrix. A deterministic protocol computing f repeatedly partitions M f into rectangles (submatrices) until every rectangle is monochromatic. Lila Fontes (University of Toronto) 2 / 12

  7. Communication complexity The model Matrix M f has entries M f [ x , y ] = f ( x , y ). A submatrix is monochromatic if f is constant on inputs in the submatrix. A deterministic protocol computing f repeatedly partitions M f into rectangles (submatrices) until every rectangle is monochromatic. Vickrey auction The 2-player Vickrey auction is defined as f : X × Y → Z where � ( x , B ) , if x ≤ y X = Y = [2 n ], Z = [2 n +1 ] and f ( x , y ) = ( y , A ) if y < x Lila Fontes (University of Toronto) 2 / 12

  8. Communication complexity The model Matrix M f has entries M f [ x , y ] = f ( x , y ). A submatrix is monochromatic if f is constant on inputs in the submatrix. A deterministic protocol computing f repeatedly partitions M f into rectangles (submatrices) until every rectangle is monochromatic. Vickrey auction The 2-player Vickrey auction is defined as f : X × Y → Z where � ( x , B ) , if x ≤ y X = Y = [2 n ], Z = [2 n +1 ] and f ( x , y ) = ( y , A ) if y < x Lila Fontes (University of Toronto) 2 / 12

  9. Communication complexity The model Matrix M f has entries M f [ x , y ] = f ( x , y ). A submatrix is monochromatic if f is constant on inputs in the submatrix. A deterministic protocol computing f repeatedly partitions M f into rectangles (submatrices) until every rectangle is monochromatic. Vickrey auction The 2-player Vickrey auction is defined as f : X × Y → Z where � ( x , B ) , if x ≤ y X = Y = [2 n ], Z = [2 n +1 ] and f ( x , y ) = ( y , A ) if y < x Lila Fontes (University of Toronto) 2 / 12

  10. Communication complexity The model Matrix M f has entries M f [ x , y ] = f ( x , y ). A submatrix is monochromatic if f is constant on inputs in the submatrix. A deterministic protocol computing f repeatedly partitions M f into rectangles (submatrices) until every rectangle is monochromatic. Vickrey auction The 2-player Vickrey auction is defined as f : X × Y → Z where � ( x , B ) , if x ≤ y X = Y = [2 n ], Z = [2 n +1 ] and f ( x , y ) = ( y , A ) if y < x Lila Fontes (University of Toronto) 2 / 12

  11. Communication complexity The model Matrix M f has entries M f [ x , y ] = f ( x , y ). A submatrix is monochromatic if f is constant on inputs in the submatrix. A deterministic protocol computing f repeatedly partitions M f into rectangles (submatrices) until every rectangle is monochromatic. Vickrey auction The 2-player Vickrey auction is defined as f : X × Y → Z where � ( x , B ) , if x ≤ y X = Y = [2 n ], Z = [2 n +1 ] and f ( x , y ) = ( y , A ) if y < x Regions (preimages) region R x , y = � ( x ′ , y ′ ) ∈ X × Y | f ( x , y ) = f ( x ′ , y ′ ) � defined by function − → Lila Fontes (University of Toronto) 2 / 12

  12. Communication complexity The model Matrix M f has entries M f [ x , y ] = f ( x , y ). A submatrix is monochromatic if f is constant on inputs in the submatrix. A deterministic protocol computing f repeatedly partitions M f into rectangles (submatrices) until every rectangle is monochromatic. Vickrey auction The 2-player Vickrey auction is defined as f : X × Y → Z where � ( x , B ) , if x ≤ y X = Y = [2 n ], Z = [2 n +1 ] and f ( x , y ) = ( y , A ) if y < x Rectangles Regions (preimages) rectangle P x , y = � ( x ′ , y ′ ) ∈ X × Y | region R x , y = � ( x ′ , y ′ ) ∈ X × Y | f ( x , y ) = f ( x ′ , y ′ ) f ( x , y ) = f ( x ′ , y ′ ) � and π ( x , y ) = π ( x ′ , y ′ ) } defined by function − → defined by protocol Lila Fontes (University of Toronto) 2 / 12

  13. Perfect privacy Privacy against eavesdroppers Can an eavesdropper learn about x and y , aside from z = f ( x , y )? Lila Fontes (University of Toronto) 3 / 12

  14. Perfect privacy Privacy against eavesdroppers Can an eavesdropper learn about x and y , aside from z = f ( x , y )? Lila Fontes (University of Toronto) 3 / 12

  15. Perfect privacy Privacy against eavesdroppers Can an eavesdropper learn about x and y , aside from z = f ( x , y )? Lila Fontes (University of Toronto) 3 / 12

  16. Perfect privacy Privacy against eavesdroppers Can an eavesdropper learn about x and y , aside from z = f ( x , y )? Lila Fontes (University of Toronto) 3 / 12

  17. Perfect privacy Privacy against eavesdroppers Can an eavesdropper learn about x and y , aside from z = f ( x , y )? Lila Fontes (University of Toronto) 3 / 12

  18. Perfect privacy Privacy against eavesdroppers Can an eavesdropper learn about x and y , aside from z = f ( x , y )? Lila Fontes (University of Toronto) 3 / 12

  19. Perfect privacy Privacy against eavesdroppers Can an eavesdropper learn about x and y , aside from z = f ( x , y )? Ascending English bidding. Lila Fontes (University of Toronto) 3 / 12

  20. Perfect privacy Perfect privacy A protocol for 2-player function f : X × Y → Z is perfectly private if every two inputs in the same region are partitioned into the same rectangle . Lila Fontes (University of Toronto) 4 / 12

  21. Perfect privacy Perfect privacy A protocol for 2-player function f : X × Y → Z is perfectly private if every two inputs in the same region are partitioned into the same rectangle . Characterizing perfect privacy (Kushilevitz ’89) The perfectly private functions of 2 inputs are fully characterized combinatorially. A private deterministic protocol for such functions is given by “decomposing” M f . Lila Fontes (University of Toronto) 4 / 12

  22. Approximate privacy Lila Fontes (University of Toronto) 5 / 12

  23. Approximate privacy Privacy approximation ratio (Feigenbaum Jaggard Schapira ’10) A protocol for f has worst-case privacy approximation ratio : | R x , y | worst-case PAR = max | P x , y | ( x , y ) | R x , y | average-case PAR = E ( x , y ) | P x , y | over distribution U Lila Fontes (University of Toronto) 5 / 12

  24. Approximate privacy Privacy approximation ratio (Feigenbaum Jaggard Schapira ’10) A protocol for f has worst-case privacy approximation ratio : | R x , y | worst-case PAR = max | P x , y | ( x , y ) | R x , y | average-case PAR = E ( x , y ) | P x , y | over distribution U Lila Fontes (University of Toronto) 5 / 12

  25. Approximate privacy Privacy approximation ratio (Feigenbaum Jaggard Schapira ’10) A protocol for f has worst-case privacy approximation ratio : | R x , y | worst-case PAR = max | P x , y | ( x , y ) | R x , y | average-case PAR = E ( x , y ) | P x , y | over distribution U Lila Fontes (University of Toronto) 5 / 12

  26. Approximate privacy Privacy approximation ratio (Feigenbaum Jaggard Schapira ’10) A protocol for f has worst-case privacy approximation ratio : | R x , y | worst-case PAR = max | P x , y | ( x , y ) | R x , y | average-case PAR = E ( x , y ) | P x , y | over distribution U worst-case PAR = 10 average-case PAR = 2 Lila Fontes (University of Toronto) 5 / 12

  27. Approximate privacy Two-player Vickrey auction Bisection protocol Lila Fontes (University of Toronto) 6 / 12

  28. Approximate privacy Two-player Vickrey auction Bisection protocol Lila Fontes (University of Toronto) 6 / 12

  29. Approximate privacy Two-player Vickrey auction Bisection protocol Lila Fontes (University of Toronto) 6 / 12

  30. Approximate privacy Two-player Vickrey auction Bisection protocol Lila Fontes (University of Toronto) 6 / 12

  31. Approximate privacy Two-player Vickrey auction Bisection protocol Lila Fontes (University of Toronto) 6 / 12

  32. Approximate privacy Two-player Vickrey auction Bisection protocol Lila Fontes (University of Toronto) 6 / 12

  33. Approximate privacy Upper bounds (Feigenbaum Jaggard Schapira ’10) English bidding bisection protocol 2 n communication cost O ( n ) 2 n worst-case PAR 1 average-case PAR 1 O (1) Lila Fontes (University of Toronto) 7 / 12

Recommend


More recommend