readytensor / rt_roberta_pii_redactor

Roberta for PII redaction

Personally identifiable information (PII) Redactor with RoBERTa

PII Redactor model using RoBERTa. The goal of this model is to remove personally identifiable information from text documents.

PII includes:

• Names
• Dates
• Emails
• Phone numbers
• Addresses
• URLs

---

Here are the highlights of this implementation:

• RoBERTa model trained using transformers library. The model is trained to identify names, addresses and dates.
• Regex logic to capture phone numbers, emails and URLs.
• FakeGenerator module to generate fake information.
• Redactor module to replace PII with fake information.

Project Structure

The following is the directory structure of the project:

• model_inputs_outputs/: This directory contains files that are either inputs to, or outputs from, the model. This directory is further divided into:
  • /inputs/: This directory contains the input .txt files to be redacted.
  • /model: This directory is used to store the model used for redaction along with the tokenizer used for tokenizing the text files.
  • /outputs/: The outputs directory will contain the output files after running the model on the input files.
• src/: This directory holds the source code for the project. It is further divided into various subdirectories:
  • config/: for configuration files for data preprocessing, model hyperparameters, paths, etc.
  • main.py: This script is used to run the model on the text files inside inputs directory.
  • utils.py: This script contains utility functions used by the other scripts.
• .gitignore: This file specifies the files and folders that should be ignored by Git.
• LICENSE: This file contains the license for the project.
• requirements.txt for the main code in the src directory.
• label2id.json This file contains label encoding for the token classes that were used to train the model.
• README.md: This file (this particular document) contains the documentation for the project, explaining how to set it up and use it.

Usage

• Place the data you want to redact in a .txt or .pdf format
• Move the files inside the /model_inputs_outputs/inputs directory
• Run the main.py script
• Get the result files from /model_inputs_outputs/outputs directory

Requirements

Dependencies for the main model implementation in src are listed in the file requirements.txt. You can install these packages by running the following command from the root of your project directory:

pip install -r requirements.txt

LICENSE

This project is provided under the Apache-2.0 License. Please see the LICENSE file for more information.

Contact Information

Repository created by Ready Tensor, Inc. Visit https://www.readytensor.ai/