mlsanityData to 7 Oct 2026

Guiding Text-to-Text Privatization by Syntax

arXiv 2306.01471 ↗First posted 2 Jun 2023cs.CL cs.CR cs.LGcited by 14

Not on the map. Why?

Abstract (arXiv)

Metric Differential Privacy is a generalization of differential privacy tailored to address the unique challenges of text-to-text privatization. By adding noise to the representation of words in the geometric space of embeddings, words are replaced with words located in the proximity of the noisy representation. Since embeddings are trained based on word co-occurrences, this mechanism ensures that substitutions stem from a common semantic context. Without considering the grammatical category of words, however, this mechanism cannot guarantee that substitutions play similar syntactic roles. We analyze the capability of text-to-text privatization to preserve the grammatical category of words after substitution and find that surrogate texts consist almost exclusively of nouns. Lacking the capability to produce surrogate texts that correlate with the structure of the sensitive texts, we encompass our analysis by transforming the privatization step into a candidate selection problem in which substitutions are directed to words with matching grammatical properties. We demonstrate a substantial improvement in the performance of downstream tasks by up to 4.66% while retaining comparative privacy guarantees.

Key facts. Cited by 14 papers in the map (data through 7 Oct 2026). Builds on 21 of its 41 references that are in the map. First posted on arXiv 2 Jun 2023.

Builds on

21 of 41 references.

References · 21

  1. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,204
  2. 2018GLUE: A Multi-Task Benchmark and Analysis Platform for Natural Language Understandingcited by 4,854
  3. 2013Efficient Estimation of Word Representations in Vector Spacecited by 4,346
  4. 2016Membership Inference Attacks against Machine Learning Modelscited by 2,221
  5. 2020Extracting Training Data from Large Language Modelscited by 2,144
  6. 2016Enriching Word Vectors with Subword Informationcited by 1,578
  7. 2017Renyi Differential Privacycited by 643
  8. 2020Information Leakage in Embedding Modelscited by 199
  9. 2019Privacy- and Utility-Preserving Textual Analysis via Calibrated Multivariate Perturbationscited by 115
  10. 2022The Limits of Word Level Differential Privacycited by 63
  11. 2018Privacy-preserving Neural Representations of Textcited by 54
  12. 2018Generalised Differential Privacy for Text Document Processingcited by 52
9 more
  1. 2019Leveraging Hierarchical Representations for Preserving Privacy and Utility in Textcited by 49
  2. 2021TEM: High Utility Metric Differential Privacy on Textcited by 43
  3. 2020A Differentially Private Text Perturbation Method Using a Regularized Mahalanobis Metriccited by 41
  4. 2014C-sanitized: a privacy model for document redaction and sanitizationcited by 27
  5. 2018SynTF: Synthetic and Differentially Private Term Frequency Vectors for Privacy-Preserving Text Miningcited by 20
  6. 2021On a Utilitarian Approach to Privacy Preserving Text Generationcited by 17
  7. 2021BRR: Preserving Privacy of Text Data Efficiently on Devicecited by 6
  8. 2019dpUGC: Learn Differentially Private Representation for User Generated Contentscited by 5
  9. 2018Author Obfuscation Using Generalised Differential Privacycited by 4

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2023: 2; 2024: 2; 2025: 7; 2026: 3.

14 papers.

Sort:

  1. 2023Privacy-Preserving In-Context Learning with Differentially Private Few-Shot Generationcited by 94
  2. 2024Data-adaptive Differentially Private Prompt Synthesis for In-Context Learningcited by 10
  3. 2025Spend Your Budget Wisely: Towards an Intelligent Distribution of the Privacy Budget in Differentially Private Text Rewritingcited by 5
  4. 2025How do we measure privacy in text? A survey of text anonymization metricscited by 5
  5. 2024Private Language Models via Truncated Laplacian Mechanismcited by 3
  6. 2025Leveraging Semantic Triples for Private Document Generation with Local Differential Privacy Guaranteescited by 3
  7. 2025Investigating User Perspectives on Differentially Private Text Privatizationcited by 1
  8. 2025The Double-edged Sword of LLM-based Data Reconstruction: Understanding and Mitigating Contextual Vulnerability in Word-level Differential Privacy Text Sanitizationcited by 1
  9. 2026A Systematic Exploration of Text Decomposition and Budget Distribution in Differentially Private Text Obfuscationcited by 1
  10. 2023Disentangling the Linguistic Competence of Privacy-Preserving BERTcited by 0
  11. 2025Stereo-LiDAR Fusion by Semi-Global Matching With Discrete Disparity-Matching Cost and Semidensificationcited by 0
  12. 2025One Size Fits All? A Modular Adaptive Sanitization Kit (MASK) for Customizable Privacy-Preserving Phone Scam Detectioncited by 0
2 more
  1. 2026Differentially-Private Text Rewriting reshapes Linguistic Stylecited by 0
  2. 2026PrivBench: A Holistic and Modular Benchmarking Platform for Evaluating Text-to-Text Privatizationcited by 0