跳到正文
研踪PaperTrail
← 研踪 PaperTrail

arXiv:2611.11467 · v3

PLDP-IDS: Personalized Local Differential Privacy for Infinite Data Stream

PLDP-IDS：面向无限数据流的个性化本地差分隐私

Leilei DuYikun HuXu ZhouYang CaoPeng ChengWangze NiKenli Li
cs.DB
首次提交 2026-11-09 00:00
版本更新 2026-11-13 01:56
雷达观测 2026-11-13 14:15
arXiv ↗PDF ↗

摘要

Existing ww-event local differential privacy (LDP) mechanisms for streaming data impose a uniform privacy window and budget, forcing all users to adopt the most stringent privacy requirement. We study personalized continual frequency estimation under LDP, where each user specifies her own window size and privacy budget, and identify two challenges. First, a single sampling size cannot balance sampling and perturbation errors for privacy groups with heterogeneous requirements. Second, borrowing estimates from groups with larger budgets reduces variance but introduces bias when privacy preferences correlate with the data. We first propose Optimal Population Selection (OPS) and two OPS-guided mechanisms, PLPD and PLPA, which use one privacy-compatible sampling size for all users. Since one size cannot fit all groups and pooling their estimates is biased under privacy--data correlation, we further develop PLPD+^+ and PLPA+^+, which combine group-aware population scheduling with bias-aware cross-group fusion while preserving personalized ww-event LDP. Experiments on real-world and synthetic streams show that, compared with the direct personalized extension of LDP-IDS, PLPA+^+ reduces the average mean squared error (AMSE) by 52.3\% and 40.8\% on average, respectively, and remains 4.3--5.1×\times more accurate than PLPA under strong privacy--data correlation.

查看 arXiv 原文 ↗阅读 PDF ↗

AI 论文解读

基于摘要 · 非全文审阅

针对无限数据流中用户隐私需求异构的问题，提出个性化 w-event 本地差分隐私下的持续频率估计机制。

正在读取账号与研究记录…

版本差异

仅对比实际观测的元数据，不代表 PDF 正文或实验变化。版本可能不连续；带删除线为移除，下划线为新增。

摘要变化

Streaming data collectionExisting under$w$-event local differential privacy (LDP) enables real-time analytics without a trusted curator. Existing w-eventmechanisms LDPfor mechanisms,streaming however,data impose a uniform privacy window and budget, onforcing all users, forcing everyone to adopt the most stringent privacy requirement. We study personalized continual frequency estimation under LDP, where each user specifies her own window size and privacy budget, and identify two challenges. First, a single sampling size cannot balance sampling and perturbation errors for privacy groups with heterogeneous requirements. Second, borrowing estimates from groups with larger budgets reduces variance but introduces bias when privacy preferences correlate with the data. We first propose Optimal Population Selection (OPS) and two OPS-guided mechanisms, PLPD and PLPA, which use one privacy-compatible sampling size for all users. WeSince thenone size cannot fit all groups and pooling their estimates is biased under privacy--data correlation, we further develop PLPD$^+$ and PLPA$^+$, which combine group-aware population scheduling with bias-aware cross-group fusion while preserving personalized $w$-event LDP. Experiments on real-world and synthetic streams show that PLPA+ reduces the, AMSEcompared ofwith the direct personalized extension of LDP-IDS, PLPA$^+$ reduces the average mean squared error (AMSE) by 52.3\% and 40.8\% on average, respectively, and remains about $4.3--5.1$\times$ more accurate than PLPA under strong privacy--data correlation.

观测到的版本

相邻记录可能跳过中间版本；首次发现 v2 及以上也会标为版本更新。

v3 · 2026-11-13 01:56 · 最新观测版本

PLDP-IDS: Personalized Local Differential Privacy for Infinite Data Stream

Existing ww-event local differential privacy (LDP) mechanisms for streaming data impose a uniform privacy window and budget, forcing all users to adopt the most stringent privacy requirement. We study personalized continual frequency estimation under LDP, where each user specifies her own window size and privacy budget, and identify two challenges. First, a single sampling size cannot balance sampling and perturbation errors for privacy groups with heterogeneous requirements. Second, borrowing estimates from groups with larger budgets reduces variance but introduces bias when privacy preferences correlate with the data. We first propose Optimal Population Selection (OPS) and two OPS-guided mechanisms, PLPD and PLPA, which use one privacy-compatible sampling size for all users. Since one size cannot fit all groups and pooling their estimates is biased under privacy--data correlation, we further develop PLPD+^+ and PLPA+^+, which combine group-aware population scheduling with bias-aware cross-group fusion while preserving personalized ww-event LDP. Experiments on real-world and synthetic streams show that, compared with the direct personalized extension of LDP-IDS, PLPA+^+ reduces the average mean squared error (AMSE) by 52.3\% and 40.8\% on average, respectively, and remains 4.3--5.1×\times more accurate than PLPA under strong privacy--data correlation.

查看此版本解读 →查看此版本原文 ↗
v2 · 2026-11-10 17:31

PLDP-IDS: Personalized Local Differential Privacy for Infinite Data Stream

Streaming data collection under local differential privacy (LDP) enables real-time analytics without a trusted curator. Existing w-event LDP mechanisms, however, impose a uniform privacy window and budget on all users, forcing everyone to adopt the most stringent requirement. We study personalized continual frequency estimation under LDP, where each user specifies her own window size and privacy budget, and identify two challenges. First, a single sampling size cannot balance sampling and perturbation errors for privacy groups with heterogeneous requirements. Second, borrowing estimates from groups with larger budgets reduces variance but introduces bias when privacy preferences correlate with the data. We first propose Optimal Population Selection (OPS) and two OPS-guided mechanisms, PLPD and PLPA. We then develop PLPD+ and PLPA+, which combine group-aware population scheduling with bias-aware cross-group fusion while preserving personalized w-event LDP. Experiments on real-world and synthetic streams show that PLPA+ reduces the AMSE of the direct personalized extension of LDP-IDS by 52.3% and 40.8% on average, respectively, and remains about 5×5\times more accurate than PLPA under strong privacy--data correlation.

查看此版本解读 →查看此版本原文 ↗
v1 · 2026-11-09 00:00

PLDP-IDS: Personalized Local Differential Privacy for Infinite Data Stream

Streaming data collection under local differential privacy (LDP) enables real-time analytics without a trusted curator. Existing ww-event LDP mechanisms, however, impose a uniform privacy window and budget on all users, forcing everyone to adopt the most stringent requirement. We study personalized continual frequency estimation under LDP, where each user specifies her own window size and privacy budget, and identify two challenges. First, a single sampling size cannot balance sampling and perturbation errors for privacy groups with heterogeneous requirements. Second, borrowing estimates from groups with larger budgets reduces variance but introduces bias when privacy preferences correlate with the data. We first propose Optimal Population Selection (OPS) and two OPS-guided mechanisms, PLPD and PLPA. We then develop PLPD+ and PLPA+, which combine group-aware population scheduling with bias-aware cross-group fusion while preserving personalized w-event LDP. Experiments on real-world and synthetic streams show that PLPA+ reduces the AMSE of the direct personalized extension of LDP-IDS by 52.3% and 40.8% on average, respectively, and remains about 5×5\times more accurate than PLPA under strong privacy--data correlation.

查看此版本解读 →查看此版本原文 ↗