erfanghorbanee / Mondrian-Multidimensional-K-Anonymity

An application of the "Mondrian Multidimensional K-Anonymity" article in Python

Mondrian-Multidimensional-K-Anonymity

An application of the "Mondrian Multidimensional K-Anonymity" article in Python.

It provides a robust anonymization technique for protecting user privacy which prevents table joining attacks on datasets while preserving data utility.

Installation

Clone the repository and set up a virtual environment:

cd Mondrian-Multidimensional-K-Anonymity

python3 -m venv venv
. venv/bin/activate

Install the requirements:

pip install -r requirements.txt

Usage

To run the Mondrian anonymization process on your data:

python3 mondrian.py --input data/adult.csv --sensitive-data class

You can also get an illustration of c-avg metric for different k values using data/adult.csv file:

python3 mondrian.py --test

Note that this is only an illustration of algorithm quality and will be generated using predefined data and the generated figure is always the same.

To see the full list of commands run:

python3 mondrian.py --help

usage: mondrian.py [-h] [--ei Explicit Identifier [Explicit Identifier ...]]
                   [--sensitive-data Sensitive Data [Sensitive Data ...]] [--k K] [--input INPUT] [--test]

An application of 'Mondrian Multidimensional K-Anonymity' article in Python

options:
  -h, --help            show this help message and exit
  --ei Explicit Identifier [Explicit Identifier ...]
                        Explicit Identifiers to be removed from dataset (example: --ei id name)
  --sensitive-data Sensitive Data [Sensitive Data ...]
                        Sensitive Data you don't want to anonymize to maintain utility (example: --sensitive-
                        data salary)
  --k K                 The k value for k-anonymity (default: 2)
  --input INPUT         Input csv file path (default: data/adult.csv)
  --test                Draws an illustration of c-avg metric for different k values using data/adult.csv file.

Resources

Article

• Mondrian Multidimensional K-Anonymity by K. LeFevre; D.J. DeWitt; R. Ramakrishnan

Dataset

• Adults database from the UC Irvine Machine Learning Repository

Contribution

Feel free to contribute in any way possible.