# Unsupervised Text Deidentification

arXiv: https://arxiv.org/abs/2210.11528 · HTML: https://mlsanity.com/paper/2210.11528/

## Key facts

- Cited by 11 papers in the map (data through 14 Oct 2026).
- Builds on 15 of its 45 references that are in the map.
- First posted on arXiv 20 Oct 2022.

## Builds on (most cited first, 15 of 15)

- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [RoBERTa: A Robustly Optimized BERT Pretraining Approach](https://mlsanity.com/paper/1907.11692/) (2019)
- [Dense Passage Retrieval for Open-Domain Question Answering](https://mlsanity.com/paper/2004.04906/) (2020)
- [When Does Label Smoothing Help?](https://mlsanity.com/paper/1906.02629/) (2019)
- [TAPAS: Weakly Supervised Table Parsing via Pre-training](https://mlsanity.com/paper/2004.02349/) (2020)
- [Large Language Models Can Be Strong Differentially Private Learners](https://mlsanity.com/paper/2110.05679/) (2021)
- [Understanding Neural Networks through Representation Erasure](https://mlsanity.com/paper/1612.08220/) (2016)
- [DeepPrivacy: A Generative Adversarial Network for Face Anonymization](https://mlsanity.com/paper/1909.04538/) (2019)
- [The Text Anonymization Benchmark (TAB): A Dedicated Corpus and Evaluation Framework for Text Anonymization](https://mlsanity.com/paper/2202.00443/) (2022)
- [CIAGAN: Conditional Identity Anonymization Generative Adversarial Networks](https://mlsanity.com/paper/2005.09544/) (2020)
- [Learning Variational Word Masks to Improve the Interpretability of Neural Text Classifiers](https://mlsanity.com/paper/2010.00667/) (2020)
- [Probabilistically Masked Language Model Capable of Autoregressive Generation in Arbitrary Word Order](https://mlsanity.com/paper/2004.11579/) (2020)
- [SynthBio: A Case Study in Human-AI Collaborative Curation of Text Datasets](https://mlsanity.com/paper/2111.06467/) (2021)
- [PHICON: Improving Generalization of Clinical Text De-identification Models via Data Augmentation](https://mlsanity.com/paper/2010.05143/) (2020)
- [Adversarial Learning of Privacy-Preserving Text Representations for De-Identification of Medical Records](https://mlsanity.com/paper/1906.05000/) (2019)

## Built on by (most cited first, 11 of 11)

- [A False Sense of Privacy: Evaluating Textual Data Sanitization Beyond Surface-level Privacy Leakage](https://mlsanity.com/paper/2504.21035/) (2025)
- [Large-scale online deanonymization with LLMs](https://mlsanity.com/paper/2602.16800/) (2026)
- [Do CoNLL-2003 Named Entity Taggers Still Work Well in 2023?](https://mlsanity.com/paper/2212.09747/) (2022)
- [More RLHF, More Trust? On The Impact of Preference Alignment On Trustworthiness](https://mlsanity.com/paper/2404.18870/) (2024)
- [DIRI: Adversarial Patient Reidentification with Large Language Models for Evaluating Clinical Text Anonymization](https://mlsanity.com/paper/2410.17035/) (2024)
- [Can Large Language Models Really Recognize Your Name?](https://mlsanity.com/paper/2505.14549/) (2025)
- [Towards Privacy-aware Mental Health AI Models: Advances, Challenges, and Opportunities](https://mlsanity.com/paper/2502.00451/) (2025)
- [Comparing Feature-based and Context-aware Approaches to PII Generalization Level Prediction](https://mlsanity.com/paper/2407.02837/) (2024)
- [Beyond Indistinguishability: Measuring Extraction Risk in LLM APIs](https://mlsanity.com/paper/2604.18697/) (2026)
- [Protecting De-identified Documents from Search-based Linkage Attacks](https://mlsanity.com/paper/2510.06383/) (2025)
- [The Conundrum of Trustworthy Research on Attacking Personally Identifiable Information Removal Techniques](https://mlsanity.com/paper/2603.08207/) (2026)

## Abstract (arXiv)

Deidentification seeks to anonymize textual data prior to distribution. Automatic deidentification primarily uses supervised named entity recognition from human-labeled data points. We propose an unsupervised deidentification method that masks words that leak personally-identifying information. The approach utilizes a specially trained reidentification model to identify individuals from redacted personal documents. Motivated by K-anonymity based privacy, we generate redactions that ensure a minimum reidentification rank for the correct profile of the document. To evaluate this approach, we consider the task of deidentifying Wikipedia Biographies, and evaluate using an adversarial reidentification metric. Compared to a set of unsupervised baselines, our approach deidentifies documents more completely while removing fewer words. Qualitatively, we see that the approach eliminates many identifying aspects that would fall outside of the common named entity based approach.