This project is an implementation of

• Differentially Private High-Dimensional Data Publication via Sampling-Based Inference
• PriView: practical differentially private release of marginal contingency tables

The Docker image is now available on Docker Hub.

Using the following commands to launch the application.

docker run -itd \ 
     -p HOST_PORT:8080 \ 
     -v USER_DATA:/opt/de-identification/static/test/ \
     -v USER_LOG:/opt/de-identification/log/ \ 
     robinlin/de-identification \
     /bin/bash

where

• HOST_PORT: The port available on your machine, e.g. 8080 or 8888
• USER_DATA: The path you place your sensitive data, so that the application can read from.
• USER_LOG: The path you could find the execution log. Be sure to substitute HOST_PORT, USER_DATA and USER_LOG with your own settings.

See here a quick start.