arXiv cs.AI· Linh Uyen Le, Ho Dinh Tri, Christian Hoang, Huy Hoang Ha·· 9 天前AI 评分27
Meddies-PII：面向临床去标识化的多语言 PII 抽取框架
Meddies-PII: A Multilingual Framework for Personally Identifiable Information Extraction in Clinical De-identification
AI 导读
Meddies-PII-Dataset 是一个覆盖 17 种语言、9 类 PII 标签的百万级合成临床文档语料，通过属性条件提示词生成，并经 13 道确定性校验门控保证结构与标注一致性。
正文 · 原文
Abstract:Clinical de-identification relies on accurately identifying personally identifiable information (PII). However, manually annotated datasets are costly to construct, while existing synthetic alternatives often provide limited details about their generation process or rely on relatively simple synthesis strategies. We introduce Meddies-PII-Dataset, a corpus of one million synthetic clinical documents spanning seventeen languages and nine PII labels. The documents are generated using attribute-conditioned prompts and validated through thirteen deterministic gates that enforce structural and annotation consistency. To evaluate the dataset's utility, we train Meddies-PII-Model, a BIOES token classifier, and compare it with existing PII extraction systems using exact-match entity-level F1. Meddies-PII-Model achieves the highest performance among the evaluated systems on all reported benchmarks, with a mean F1 of 0.827 across fifteen external benchmarks, compared with 0.658 for the strongest baseline. Upon acceptance, we will publicly release the dataset, benchmark suite, model, generation framework, and evaluation code to support research on multilingual clinical de-identification.
| Subjects: | Computation and Language (cs.CL); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2609.12544 [cs.CL] |
| (or arXiv:2609.12544v2 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2609.12544 arXiv-issued DOI via DataCite |
Submission history
From: Christian Hoang [view email]
[v1]
Fri, 11 Sep 2026 07:53:52 UTC (1,581 KB)
[v2]
Thu, 8 Oct 2026 13:00:49 UTC (1,581 KB)
来源：arXiv cs.AI · arxiv.org