Structured hierarchical method improves private data insights for analytics

SHIP: Structured and Hierarchical Insight with Privacy

Computation and LanguageCryptography and SecurityMachine Learning

Summary

Extracting useful information from text data while keeping people's privacy safe is a hard problem. The authors introduce SHIP, a new way to organize and represent private information so that many types of analytical questions can be answered more accurately without revealing sensitive details. Their method builds a hierarchy of data categories and releases a synthetic, privacy-protected summary table, which helps researchers better understand the data. In tests, SHIP performed better than other methods, though some questions remain challenging.

What this means in practice

  • For hospital data teams: Generate privacy-protected synthetic health records to enable safe analysis of multi-attribute patient data.
  • For marketing analytics teams: Create synthetic customer datasets that support detailed queries while preserving consumer privacy for targeted marketing insights.

Authors

Kai Chen, Guisi Ni, Albert Cheu, Kassem Fawaz, Yuzheng Hu, Ryan McKenna, Peter Kairouz

Abstract

Extracting insights from textual data while protecting user privacy is increasingly important. Existing methods produce private insights in diverse forms for different analytical purposes, making their utility difficult to assess and compare. In this work, we formulate differentially private (DP) insights as reusable representations for answering analytical queries, providing a common basis for utility assessment. Answering these queries, however, requires fine-grained details and cross-attribute statistics that existing representations may omit. To capture this information, we introduce Structured and Hierarchical Insight with Privacy (SHIP), which automatically maps queries to attributes, builds a hierarchical hybrid taxonomy, and releases a DP synthetic table for further analysis. We also develop an evaluation protocol based primarily on QA performance and structured fidelity. Our experiments demonstrate SHIP's strong performance on both measures relative to existing private insight baselines, while identifying queries that remain challenging for structured representations.