Empowering AI to Protect User Privacy


Stony Brook Computer Science PhD student alum Aria Rezaei and Professor Jie Gao have been awarded the Best Paper Award at EWSN 2021 for "Application-driven Privacy-preserving Data Publishing with Correlated Attributes."

The International Conference on Embedded Wireless Systems and Networks (EWSN) is an international conference focusing on the latest research related to embedded systems, wireless networking, the ‘Internet of Things’ and Cyber-Physical Systems. EWSN ‘21 featured the theme of “Embedded AI,” reflecting the recent technological advances at the crossroads of embedded systems and the vision of revolutionizing AI capabilities from the cloud to physical devices.

Rezaei and Gao’s research is focused on users' privacy protection in Machine Learning applications.  Privacy is an issue that is relevant to everyone in our society, with their research examining the issue of protecting sensitive information that can be learned from raw data. Rezaei and Gao seek to develop methods to protect user privacy and at the same time allow tech applications to be successful in their goals.

In an ideal world, a user would be fully in charge of their personal information, deciding when, where and for what purpose that information is used. It is critical to balance the protection of user privacy and societal benefits. Healthcare data, for example, has many privacy issues. Smart building applications that engage in occupancy counting can be useful for efficient energy management, but there is the risk that users are identified from the collected signals and thus the data can be used for unintended purposes. How to learn useful group-level knowledge while protecting individual identity and privacy is an ongoing challenge that Rezaei  and Gao address.

Congratulations to Rezaei and Gao on this well deserved recognition at EWSN 2021 for their timely and impactful research.


      Jie Gao                                                                          Aria Rezaei



- Andrew Solar-Greco