arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2109.03175cs.CLcs.CR
收藏

当差分隐私遇到NLP:魔鬼就在细节中

When differential privacy meets NLP: The devil is in the detail

Technical University of Darmstadt(达姆施塔特工业大学)

Ivan Habernal

摘要:

差分隐私提供了一个正式的方法来保护个人隐私。差分隐私在各种场景中的应用,例如保护用户的原始话语,必须满足某些数学性质。我们的贡献是对ADePT的正式分析,ADePT是一种用于文本重写的差分私有自动编码器(Krishna等人,2021)。ADePT在下游任务上取得了令人鼓舞的成果,同时提供了严格的隐私保证。我们的证据表明,ADePT是没有差异的私人,从而使实验结果未经证实。我们还量化了其私有机制中的错误的影响,表明在非常小的编码器维度的乐观情况下,真实的灵敏度至少高出6倍,并且未私有化的话语量可以轻松达到整个数据集的100%。我们的意图既不是批评作者,也不是批评同行评审过程,而是指出,如果NLP中的差分隐私应用依赖于正式的保证,这些应该被全面概述并进行详细审查。

英文摘要:

Differential privacy provides a formal approach to privacy of individuals. Applications of differential privacy in various scenarios, such as protecting users' original utterances, must satisfy certain mathematical properties. Our contribution is a formal analysis of ADePT, a differentially private auto-encoder for text rewriting (Krishna et al, 2021). ADePT achieves promising results on downstream tasks while providing tight privacy guarantees. Our proof reveals that ADePT is not differentially private, thus rendering the experimental results unsubstantiated. We also quantify the impact of the error in its private mechanism, showing that the true sensitivity is higher by at least factor 6 in an optimistic case of a very small encoder's dimension and that the amount of utterances that are not privatized could easily reach 100% of the entire dataset. Our intention is neither to criticize the authors, nor the peer-reviewing process, but rather point out that if differential privacy applications in NLP rely on formal guarantees, these should be outlined in full and put under detailed scrutiny.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑