Controlled Generation for Private Synthetic Text

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Zhao, Zihao, Field, Anjalie
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866908568380243968
author Zhao, Zihao
Field, Anjalie
author_facet Zhao, Zihao
Field, Anjalie
contents Text anonymization is essential for responsibly developing and deploying AI in high-stakes domains such as healthcare, social services, and law. In this work, we propose a novel methodology for privacy-preserving synthetic text generation that leverages the principles of de-identification and the Hiding In Plain Sight (HIPS) theory. Our approach introduces entity-aware control codes to guide controllable generation using either in-context learning (ICL) or prefix tuning. The ICL variant ensures privacy levels consistent with the underlying de-identification system, while the prefix tuning variant incorporates a custom masking strategy and loss function to support scalable, high-quality generation. Experiments on legal and clinical datasets demonstrate that our method achieves a strong balance between privacy protection and utility, offering a practical and effective solution for synthetic text generation in sensitive domains.
format Preprint
id arxiv_https___arxiv_org_abs_2509_25729
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Controlled Generation for Private Synthetic Text
Zhao, Zihao
Field, Anjalie
Computation and Language
Artificial Intelligence
Machine Learning
Text anonymization is essential for responsibly developing and deploying AI in high-stakes domains such as healthcare, social services, and law. In this work, we propose a novel methodology for privacy-preserving synthetic text generation that leverages the principles of de-identification and the Hiding In Plain Sight (HIPS) theory. Our approach introduces entity-aware control codes to guide controllable generation using either in-context learning (ICL) or prefix tuning. The ICL variant ensures privacy levels consistent with the underlying de-identification system, while the prefix tuning variant incorporates a custom masking strategy and loss function to support scalable, high-quality generation. Experiments on legal and clinical datasets demonstrate that our method achieves a strong balance between privacy protection and utility, offering a practical and effective solution for synthetic text generation in sensitive domains.
title Controlled Generation for Private Synthetic Text
topic Computation and Language
Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2509.25729