On the Impact of Noise in Differentially Private Text Rewriting

TL;DR

This paper investigates how noise affects the utility and privacy in differential privacy-based text rewriting, introducing a new sentence infilling method and comparing it with non-DP approaches.

Contribution

It introduces a novel sentence infilling privatization technique and empirically analyzes the impact of noise in DP text rewriting compared to non-DP methods.

Findings

Non-DP methods better preserve utility.

DP methods offer stronger privacy protections.

Noise significantly impacts the effectiveness of DP in NLP.

Abstract

The field of text privatization often leverages the notion of $\textit{Differential Privacy}$ (DP) to provide formal guarantees in the rewriting or obfuscation of sensitive textual data. A common and nearly ubiquitous form of DP application necessitates the addition of calibrated noise to vector representations of text, either at the data- or model-level, which is governed by the privacy parameter $\varepsilon$. However, noise addition almost undoubtedly leads to considerable utility loss, thereby highlighting one major drawback of DP in NLP. In this work, we introduce a new sentence infilling privatization technique, and we use this method to explore the effect of noise in DP text rewriting. We empirically demonstrate that non-DP privatization techniques excel in utility preservation and can find an acceptable empirical privacy-utility trade-off, yet cannot outperform DP methods in empirical privacy protections. Our results highlight the significant impact of noise in current DP rewriting mechanisms, leading to a discussion of the merits and challenges of DP in NLP, as well as the opportunities that non-DP methods present.