信息安全研究2026，Vol.12Issue(3)：220-227,8.DOI:10.12379/j.issn.2096-1057.2026.03.03
基于梯度方向筛选的差分隐私文本合成
Differentially Private Text Synthesis Based on Gradient Direction Filtering
摘要
Abstract
Deep learning models enhance performance by memorizing training data,but this also poses a risk of training data leakage.Differential privacy,as a mainstream privacy protection method,effectively mitigates this risk.However,existing differentially private data synthesis approaches suffer from slow model convergence and low data usability.To address these issues,we propose the TVDPSGD-LM_D framework.This approach introduces TVDPSGD,a threshold-validated differentially private optimization algorithm that incorporates a validation mechanism to filter gradient directions during differentially private model training.By discarding ineffective updates,this approach accelerates model convergence.TVDPSGD-LM embeds TVDPSGD into a language generation model to synthesize labeled text datasets that maintain statistical similarity to the original data.Additionally,a pretrained classifier is used to filter the generated text,removing samples where the classification results do not match the assigned labels,thereby improving the quality of the synthetic dataset.Experimental results on public datasets demonstrate that the proposed method preserves data privacy while achieving a classification accuracy of 89.4％on the processed synthetic dataset.关键词
差分隐私/梯度方向筛选/对比筛选/文本合成/条件控制码Key words
differential privacy/gradient direction filtering/contrastive filtering/text synthesis/conditional control code分类
信息技术与安全科学引用本文复制引用
李莉,赵霖露,郭国疆,金剑炜,段晓毅..基于梯度方向筛选的差分隐私文本合成[J].信息安全研究,2026,12(3):220-227,8.基金项目
中央高校基本科研业务费资金项目(3282024006,3282024057) （3282024006,3282024057）
