Research
- Privacy-aware data systems
- Private distributed ledgers/ blockchains
- Secure and Trustworthy AI/ML systems
Privacy-aware data systems
A persistent issue in today’s society is that organizations handling private or sensitive data, i.e. health records, banking transactions, defense industry and military data etc, strive to engage in collaborative data analysis, but they are restricted from sharing it.
Our aspiration is to architect pragmatic data systems to facilitate the practice of collaborative analytics over confidential data. To accomplish this aim, we delve into the synergistic relationship among differential privacy, cryptography, and trusted hardwares by establishing innovative algorithms and system designs that articulate the interconnection between these spheres, taking into account both theoretical perspectives and pragmatic approaches to systems building.
Featured publications
- IncShrink: Architecting Efficient Outsourced Databases using Incremental MPC and Differential Privacy (SIGMOD), 2022
- DP-Sync: Hiding Update Patterns in Secure Outsourced Databases with Differential Privacy (SIGMOD), 2021
- Crypt$\epsilon$:Crypto-Assisted Differential Privacy on Untrusted Servers (SIGMOD), 2020
Private distributed ledgers/ blockchains
The permissionless nature is widely recognized as a key innovation in modern blockchain technologies. However, this inherently open arrangement inevitably raises numerous privacy issues, for instance, everyone can observe each other’s transactions. To mitigate these issues, private blockchains like Zcash have been proposed. These blockchains conceal transaction details using cryptographic primitives and employ zero-knowledge proofs to validate transactions, thereby preserving blockchain functionality.
My interests in this area primarily include: (i)Fundamentally understand the privacy implications of existing private blockchains; (ii) Explore principled techniques for designing efficient and robust private blockchains.
Featured publications
Check out our latest posts on Decentralized Thoughts, part1, part2, and our talks on Zcon4 and CESC23
Secure and Trustworthy AI/ML systems
My current research focuses on novel algorithms and hardware co-design for accelerating privacy-preserving machine learning, aiming to facilitate the practical deployment of PPML across various industries that interact with sensitive data, such as healthcare, biomedicine, banking, finance, etc.
Featured publications
- AQ2PNN: Enabling Two-party Privacy-Preserving Deep Neural Network Inference with Adaptive Quantization(MICRO), 2023
- AutoReP: Automatic ReLU Replacement for Fast Private Network Inference (ICCV), 2023
- Against Membership Inference Attack: Pruning is All You Need (IJCAI), 2021