arXiv cs.AI· Hardeep Bassi, Hui Ma, Lu\'is Seabra, Koowum Joshi, Daniel Pereira, Averell Gatton·· 5 小时前AI 评分26
Protegrity-PII 技术报告：新型 PII 检测模型
Protegrity-PII Technical Report
AI 导读
Protegrity-PII 是一个面向结构化与非结构化文本的 PII 检测模型，在作者提出的新基准 PII-SCOPE 上 typed recall 达 0.910、precision 为 0.902。
正文 · 原文
Abstract:We present Protegrity-PII, a new PII detection model for structured and unstructured text. Protegrity-PII achieves a typed recall of 0.910 with a precision of 0.902 on a new benchmark we propose, PII-SCOPE. Our model processes text in 0.49 seconds per million characters at \$0.0005 per million characters on a single RTX 6000 using batched inference. Due in part to the expansive nature of the entity type taxonomy used in training, Protegrity-PII achieves superior out-of-distribution generalization when benchmarked against a selection of state of the art models and datasets.
| Subjects: | Computation and Language (cs.CL); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.30518 [cs.CL] |
| (or arXiv:2610.30518v2 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2610.30518 arXiv-issued DOI via DataCite |
Submission history
From: Hardeep Bassi [view email]
[v1]
Wed, 21 Oct 2026 22:56:18 UTC (1,592 KB)
[v2]
Wed, 28 Oct 2026 02:59:21 UTC (1,594 KB)
来源：arXiv cs.AI · arxiv.org