公开(公告)号：US12001588B2

公开(公告)日：2024-06-04

申请号：US17086244

申请日：2020-10-30

Applicant: SAP SE

Inventor： Daniel Bernau ,  Philip-William Grassal ,  Hannah Keller ,  Martin Haerterich

IPC: G06F21/62 ,  G06F17/18 ,  G06F18/214 ,  G06N20/00

CPC classification number: G06F21/6254 ,  G06F17/18 ,  G06F18/2148 ,  G06N20/00

Abstract: Data is received that specifies a bound for an adversarial posterior belief ρc that corresponds to a likelihood to re-identify data points from the dataset based on a differentially private function output. Privacy parameters ε, δ are then calculated based on the received data that govern a differential privacy (DP) algorithm to be applied to a function to be evaluated over a dataset. The calculating is based on a ratio of probabilities distributions of different observations, which are bound by the posterior belief ρc as applied to a dataset. The calculated privacy parameters are then used to apply the DP algorithm to the function over the dataset. Related apparatus, systems, techniques and articles are also described.

公开(公告)号：US20220138348A1

公开(公告)日：2022-05-05

申请号：US17086244

申请日：2020-10-30

Applicant: SAP SE

Inventor： Daniel Bernau ,  Philip-William Grassal ,  Hannah Keller ,  Martin Haerterich

IPC: G06F21/62 ,  G06F17/18 ,  G06K9/62 ,  G06N20/00

Abstract: Data is received that specifies a bound for an adversarial posterior belief ρc that corresponds to a likelihood to re-identify data points from the dataset based on a differentially private function output. Privacy parameters ε, δ are then calculated based on the received data that govern a differential privacy (DP) algorithm to be applied to a function to be evaluated over a dataset. The calculating is based on a ratio of probabilities distributions of different observations, which are bound by the posterior belief ρc as applied to a dataset. The calculated privacy parameters are then used to apply the DP algorithm to the function over the dataset. Related apparatus, systems, techniques and articles are also described.

公开(公告)号：US12147577B2

公开(公告)日：2024-11-19

申请号：US18581254

申请日：2024-02-19

Applicant: SAP SE

Inventor： Daniel Bernau ,  Philip-William Grassal ,  Hannah Keller ,  Martin Haerterich

IPC: G06F21/62 ,  G06F17/18 ,  G06F18/214 ,  G06N20/00

Abstract: Data is received that specifies a bound for an adversarial posterior belief ρc that corresponds to a likelihood to re-identify data points from the dataset based on a differentially private function output. Privacy parameters ε, δ are then calculated based on the received data that govern a differential privacy (DP) algorithm to be applied to a function to be evaluated over a dataset. The calculating is based on a ratio of probabilities distributions of different observations, which are bound by the posterior belief ρc as applied to a dataset. The calculated privacy parameters are then used to apply the DP algorithm to the function over the dataset. Related apparatus, systems, techniques and articles are also described.

公开(公告)号：US20250036811A1

公开(公告)日：2025-01-30

申请号：US18904462

申请日：2024-10-02

Applicant: SAP SE

Inventor： Daniel Bernau ,  Philip-William Grassal ,  Hannah Keller ,  Martin Haerterich

IPC: G06F21/62 ,  G06F17/18 ,  G06F18/214 ,  G06N20/00

Abstract: Data is received that specifies a bound for an adversarial posterior belief pc that corresponds to a likelihood to re-identify data points from the dataset based on a differentially private function output. Privacy parameters ε, δ are then calculated based on the received data that govern a differential privacy (DP) algorithm to be applied to a function to be evaluated over a dataset. The calculating is based on a ratio of probabilities distributions of different observations, which are bound by the posterior belief pc as applied to a dataset. The calculated privacy parameters are then used to apply the DP algorithm to the function over the dataset. Related apparatus, systems, techniques and articles are also described.

公开(公告)号：US20240211635A1

公开(公告)日：2024-06-27

申请号：US18581254

申请日：2024-02-19

Applicant: SAP SE

Inventor： Daniel Bernau ,  Philip-William Grassal ,  Hannah Keller ,  Martin Haerterich

IPC: G06F21/62 ,  G06F17/18 ,  G06F18/214 ,  G06N20/00

CPC classification number: G06F21/6254 ,  G06F17/18 ,  G06F18/2148 ,  G06N20/00

Abstract: Data is received that specifies a bound for an adversarial posterior belief ρc that corresponds to a likelihood to re-identify data points from the dataset based on a differentially private function output. Privacy parameters ε, δ are then calculated based on the received data that govern a differential privacy (DP) algorithm to be applied to a function to be evaluated over a dataset. The calculating is based on a ratio of probabilities distributions of different observations, which are bound by the posterior belief ρc as applied to a dataset. The calculated privacy parameters are then used to apply the DP algorithm to the function over the dataset. Related apparatus, systems, techniques and articles are also described.