SeqPATE: Differentially Private Text Generation via Knowledge Distillation
SeqPATE: Differentially Private Text Generation via Knowledge Distillation
复制标题
DOI：
--
复制
发表时间：
2022
期刊：
影响因子：
--
通讯作者：
Zhiliang Tian;Ying Zhao;Ziyue Huang;Yu-Xiang Wang;N. Zhang;He He-He
中科院分区：
文献类型：
--
作者：
Zhiliang Tian;Ying Zhao;Ziyue Huang;Yu-Xiang Wang;N. Zhang;He He-He
保护用户数据的隐私对于文本生成模型至关重要，该模型在生成过程中可能会泄露敏感信息。差分私有(DP)学习方法提供了避免从模型输出中识别训练样本的存在的保证。PATE算法是最近提出的一种DP学习算法，在训练样本上具有很高的实用性和很强的隐私保护。但是，文本生成模型在较大的输出空间中按顺序输出令牌；经典的Pate算法不是为此设置定制的。此外，Pate可以很好地保护样本级隐私，但不是用来保护样本中的短语的。在本文中，我们提出了SeqPATE，它是Pate在文本生成中的一种扩展，它保护了个人训练样本和训练数据中敏感短语的隐私。为了使Pate适应文本生成，我们生成伪上下文，并将序列生成问题归结为下一个单词预测问题。为了处理较大的输出空间，我们提出了一种动态缩减输出空间的候选过滤策略，并对Pate的教师聚合进行了细化，避免了由于投票给大量的候选而导致的低一致性。为了进一步减少隐私损失，我们使用知识蒸馏来减少教师查询的数量。实验验证了SeqPATE在保护训练样本和敏感短语方面的有效性。
Protecting the privacy of user data is crucial for text generation models, which can leak sensitive information during generation. Differentially private (DP) learning methods provide guarantees against identifying the existence of a training sample from model outputs. PATE is a recent DP learning algorithm that achieves high utility with strong privacy protection on training samples. However, text generation models output tokens sequentially in a large output space; the classic PATE algorithm is not customized for this setting. Furthermore, PATE works well to protect sample-level privacy, but is not designed to protect phrases in samples. In this paper, we propose SeqPATE, an extension of PATE to text generation that protects the privacy of individual training samples and sensitive phrases in training data. To adapt PATE to text generation, we generate pseudo-contexts and reduce the sequence generation problem to a next-word prediction problem. To handle the large output space, we propose a candidate filtering strategy to dynamically reduce the output space, and refine the teacher aggregation of PATE to avoid low agreement due to voting for a large number of candidates. To further reduce privacy losses, we use knowledge distillation to reduce the number of teacher queries. The experiments verify the effectiveness of SeqPATE in protecting both training samples and sensitive phrases.