Improve Anonymization Practices

This repository contains the implementation of An Approach to Improve Anonymization Practices in Educational Data Mining.

Description

This project implemented a way to guide the k-anonymization process toward strategies that anonymize the least important information more than more important information for downstream machine learning prediction tasks.

Getting Started

Dependencies

• sklearn
• pandas
• numpy
• pyarxaas
• docker

Executing Program

A walkthrough of our method can be found in 'base-kanon-alldata_.ipynb'. The datasets can be found through the citations in the paper.

This open source software is provided under the MIT License. We will continue to monitor and use GitHub as our version control and issue manager.