语言建模中的选择性差分隐私
Selective Differential Privacy for Language Modeling
摘要：
随着语言模型应用的日益广泛，保护语言模型的隐私信息不被泄露已变得至关重要。以前的工作试图通过训练具有差分隐私保证的基于RNN的语言模型来应对这一挑战。然而，将经典的差分隐私应用于语言模型会导致模型性能不佳，因为底层的隐私概念过于悲观，并为数据中的所有令牌提供无差别保护。考虑到自然语言中的隐私信息是稀疏的（例如，电子邮件的大部分可能不携带个人身份信息），我们提出了一个新的隐私概念，选择性差分隐私，对数据的敏感部分提供严格的隐私保证，以提高模型的实用性。为了实现这样一个新的概念，我们开发了一个相应的隐私机制，选择性DPSGD，基于RNN的语言模型。除了语言建模之外，我们还将该方法应用于一个更具体的应用--对话系统。语言建模和对话系统构建的实验表明，所提出的隐私保护机制实现了更好的效用，同时保持在各种隐私攻击的安全性相比，基线。数据和代码发布在https://github.com/wyshi/lm_privacy上，以方便未来的研究。
英文摘要：
With the increasing applications of language models, it has become crucial to protect these models from leaking private information. Previous work has attempted to tackle this challenge by training RNN-based language models with differential privacy guarantees. However, applying classical differential privacy to language models leads to poor model performance as the underlying privacy notion is over-pessimistic and provides undifferentiated protection for all tokens in the data. Given that the private information in natural language is sparse (for example, the bulk of an email might not carry personally identifiable information), we propose a new privacy notion, selective differential privacy, to provide rigorous privacy guarantees on the sensitive portion of the data to improve model utility. To realize such a new notion, we develop a corresponding privacy mechanism, Selective-DPSGD, for RNN-based language models. Besides language modeling, we also apply the method to a more concrete application--dialog systems. Experiments on both language modeling and dialog system building show that the proposed privacy-preserving mechanism achieves better utilities while remaining safe under various privacy attacks compared to the baselines. The data and code are released at https://github.com/wyshi/lm_privacy to facilitate future research .
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。
