bert-small-pii-detection

gravitee-io/bert-small-pii-detection 🚀

A more accurate PII detector fine-tuned from prajjwal1/bert-small on the datasets described in metatada.

About the dataset:

We combined various datasets in order to cover wide range of document formats like:

1. JSON,
2. HTML,
3. XML,
4. SQL
5. Documents

Label Set

AGE, COORDINATE, CREDIT_CARD, DATE_TIME, EMAIL_ADDRESS, FINANCIAL, IBAN_CODE, IMEI,
IP_ADDRESS, LOCATION, MAC_ADDRESS, NRP, ORGANIZATION, PASSWORD, PERSON, PHONE_NUMBER,
TITLE, URL, US_BANK_NUMBER, US_DRIVER_LICENSE, US_ITIN, US_LICENSE_PLATE, US_PASSPORT, US_SSN

How to Use

Quick start (pipeline)

from transformers import AutoTokenizer, AutoModelForTokenClassification, pipeline

repo = "gravitee-io/bert-small-pii-detection"
tok = AutoTokenizer.from_pretrained(repo)
model = AutoModelForTokenClassification.from_pretrained(repo)

pipe = pipeline("token-classification", model=model, tokenizer=tok, aggregation_strategy="simple")
text = ""
pipe(text)

Evaluation

Metric: precision / recall / F1 per entity, micro/macro averages

Intended Uses & Limitations

Use this model for:

• Low resource environmens
• Redacting PII in customer support logs, dev/test environments, API traces and articles
• Real-time hints in form fields or data entry systems

Limitations:

• English-focused; other languages will degrade
• Domain drift is real: audit on your own data

---

Citation

If you use the model, please consider citing the papers:

@misc{bhargava2021generalization,
      title={Generalization in NLI: Ways (Not) To Go Beyond Simple Heuristics},
      author={Prajjwal Bhargava and Aleksandr Drozd and Anna Rogers},
      year={2021},
      eprint={2110.01518},
      archivePrefix={arXiv},
      primaryClass={cs.CL}
}

@article{DBLP:journals/corr/abs-1908-08962,
  author    = {Iulia Turc and
               Ming{-}Wei Chang and
               Kenton Lee and
               Kristina Toutanova},
  title     = {Well-Read Students Learn Better: The Impact of Student Initialization
               on Knowledge Distillation},
  journal   = {CoRR},
  volume    = {abs/1908.08962},
  year      = {2019},
  url       = {http://arxiv.org/abs/1908.08962},
  eprinttype = {arXiv},
  eprint    = {1908.08962},
  timestamp = {Thu, 29 Aug 2019 16:32:34 +0200},
  biburl    = {https://dblp.org/rec/journals/corr/abs-1908-08962.bib},
  bibsource = {dblp computer science bibliography, https://dblp.org}
}

@online{WinNT,
  author = {Benjamin Kilimnik},
  title = {{Privy} Synthetic PII Protocol Trace Dataset},
  year = 2022,
  url = {https://huggingface.co/datasets/beki/privy},
}

@online{gretel2023,
  author = {Gretel.ai},
  title = {{Synthetic PII Finance Multilingual Dataset}},
  year = 2023,
  url = {https://huggingface.co/datasets/gretelai/synthetic_pii_finance_multilingual},
}

@inproceedings{tjong-kim-sang-de-meulder-2003-introduction,
    title = "Introduction to the CoNLL-2003 Shared Task: Language-Independent Named Entity Recognition",
    author = "Tjong Kim Sang, Erik F. and De Meulder, Fien",
    booktitle = "Proceedings of the Seventh Conference on Natural Language Learning at HLT-NAACL 2003",
    year = "2003",
    url = "https://aclanthology.org/W03-0419",
}
}