# PHICON: Improving Generalization of Clinical Text De-identification Models via Data Augmentation

arXiv: https://arxiv.org/abs/2010.05143 · HTML: https://mlsanity.com/paper/2010.05143/

## Key facts

- Cited by 14 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 11 Oct 2020.

## Built on by (most cited first, 14 of 14)

- [Reducing Privacy Risks in Online Self-Disclosures with Language Models](https://mlsanity.com/paper/2311.09538/) (2023)
- [FlipDA: Effective and Robust Data Augmentation for Few-Shot Learning](https://mlsanity.com/paper/2108.06332/) (2021)
- [Unsupervised Text Deidentification](https://mlsanity.com/paper/2210.11528/) (2022)
- [MAGneT: Coordinated Multi-Agent Generation of Synthetic Multi-Turn Mental Health Counseling Sessions](https://mlsanity.com/paper/2509.04183/) (2025)
- [GAMA: A General Anonymizing Multi-Agent System for Privacy Preservation Enhanced by Domain Rules and Disproof Mechanism](https://mlsanity.com/paper/2509.10018/) (2025)
- [How Private are Language Models in Abstractive Summarization?](https://mlsanity.com/paper/2412.12040/) (2024)
- [Towards Privacy-aware Mental Health AI Models: Advances, Challenges, and Opportunities](https://mlsanity.com/paper/2502.00451/) (2025)
- [Graph2Counsel: Clinically Grounded Synthetic Counseling Dialogue Generation from Client Psychological Graphs](https://mlsanity.com/paper/2604.20382/) (2026)
- [A Unified Knowledge Graph Augmentation Service for Boosting Domain-specific NLP Tasks](https://mlsanity.com/paper/2212.05251/) (2022)
- [Out-of-Distribution Generalization in Text Classification: Past, Present, and Future](https://mlsanity.com/paper/2305.14104/) (2023)
- [EPEE: Towards Efficient and Effective Foundation Models in Biomedicine](https://mlsanity.com/paper/2503.02053/) (2025)
- [Thunder-DeID: Accurate and Efficient De-identification Framework for Korean Court Judgments](https://mlsanity.com/paper/2506.15266/) (2025)
- [EpiScreen: Early Epilepsy Detection from Electronic Health Records with Large Language Models](https://mlsanity.com/paper/2603.28698/) (2026)
- [PersianAnonymizer: Evaluating LLM-Labeled Training for Efficient NER-based Anonymization in Persian](https://mlsanity.com/paper/2609.00958/) (2026)

## Abstract (arXiv)

De-identification is the task of identifying protected health information (PHI) in the clinical text. Existing neural de-identification models often fail to generalize to a new dataset. We propose a simple yet effective data augmentation method PHICON to alleviate the generalization issue. PHICON consists of PHI augmentation and Context augmentation, which creates augmented training corpora by replacing PHI entities with named-entities sampled from external sources, and by changing background context with synonym replacement or random word insertion, respectively. Experimental results on the i2b2 2006 and 2014 de-identification challenge datasets show that PHICON can help three selected de-identification models boost F1-score (by at most 8.6%) on cross-dataset test setting. We also discuss how much augmentation to use and how each augmentation method influences the performance.