This repository contains the code and implementation for the paper Data Privacy in Trigger-Action Systems.
eTAP is a privacy-enhancing trigger-action platform that executes trigger-compute-action rules without accessing users’ private data in plaintext. It utilizes a tailored garbled circuit protocol to provide data confidentiality and execution integrity.
For each party in the trigger-action system, we provide an example Python implementation that follows the eTAP protocols:
- Trigger-action platform: a Flask server that executes the rule uploaded by the client.
- Trigger service: a Flask server that sends encrypted data to the trigger-action platform.
- Action service: a Flask server that decrypts the data received from the trigger-action platform.
- Client: a Python script that takes a rule description file as input and uploads the information required to execute the rule to the above three servers.
In addition, we provide a Python module etaplib that allows developers of trigger and action services to upgrade their existing Flask API routes to be compatible with eTAP.
Please refer to the README in their own directories for detailed instructions on installation and usage.
@inproceedings{,
title = {Data Privacy in Trigger-Action Systems},
author = {Chen, Yunang and Chowdhury, Amrita Roy and Wang, Ruizhe and Sabelfeld, Andrei and Chatterjee, Rahul and Fernandes, Earlence},
booktitle = {},
year = {2021}
}