Pritesh-2711/pii-bench

---
license: apache-2.0
pretty_name: PIIBench
task_categories:
- token-classification
language:
- en
tags:
- pii
- ner
- privacy
- benchmark
size_categories:
- 1M<n<10M
---

# PIIBench

## Description
PIIBench is a unified benchmark dataset for PII detection across multiple domains.

## Paper
- arXiv: http://arxiv.org/abs/2604.15776

## Dataset Summary
- Total records: ~1.39M
- Entity types: 48
- Format: BIO tagging

## Structure
Each example contains:
- `tokens`: list of tokens
- `labels`: BIO labels
- `source`: original data source of the sample

## Splits
- `train.jsonl`
- `validation.jsonl`
- `test.jsonl`

## Source
Ten datasets are downloaded from Hugging Face and consolidated into a unified BIO-tagged format:

| Dataset | Rows | Domain |
|---|---:|---|
| ai4privacy/pii-masking-400k | ~400k | General, 63 PII classes |
| ai4privacy/pii-masking-300k | ~300k | General + Finance (FinPII-80k) |
| gretelai/synthetic_pii_finance_multilingual | ~56k | Finance (100 doc types) |
| nvidia/Nemotron-PII | ~100k | General (50+ industries) |
| wikiann (en) | ~20k | Wikipedia, PER/ORG/LOC only |
| Babelscape/multinerd (en) | varies | Wikipedia + news, 15 types |
| DFKI-SLT/few-nerd | ~188k | Wikipedia, 66 fine-grained types |
| conll2003 | ~14k | News (Reuters), 4 types |
| nlpaueb/finer-139 | ~1.1M | Finance (SEC filings), 139 XBRL tags |
| Isotonic/pii-masking-200k | ~200k | General, 54 PII classes |

`finer-139` is capped at 150k records during data preparation. Entity types with fewer than 500 B-mentions globally are collapsed to `O`.

## License
This dataset is derived from multiple sources. Users must comply with the original dataset licenses of the constituent datasets.

## Citation

```bibtex
@article{jha2026piibench,
title={PIIBench: A Unified Multi-Source Benchmark Corpus for PII Detection},
author={Jha, Pritesh},
year={2026}
}