Improving privacy-preserving computation

For example, medical records are often distributed across a wide timeline and spectrum of agencies and institutions. It can be hard for practitioners and others to share and analyze this sensitive information – and thus determine the most effective treatments, or find crucial correlations, such as between a genetic condition and specific genes.

Marina Blanton, an associate professor in the Department of Computer Science and Engineering, seeks to address challenges such as these. Her research on privacy-preserving computation focuses on developing novel cryptographic techniques that can extract knowledge and make correlations from a wide variety of sensitive data sets, while complying with legal guidelines and other provisions. Her techniques ensure that even the computation participants, often medical staff or data analysts, are unable to make inferences about the private data beyond the result that they obtain.

Funded by Google through its Faculty Research Award program, her project, "Efficient Tools for Privacy-Preserving Data Analysis," has the potential to impact a wide variety of organizations in health, education, private and government sectors.

“Improvements in the speed of these techniques are especially important in our era of big data, where the volumes of data we process are only expected to grow. Wider adoption of this technology could have an impact on different aspects of our lives ranging from facilitation of biomedical and genomic research to improved collaboration across organizational boundaries,” said Blanton.

“Google already employs privacy-preserving computation for determining how much revenue ads placed with Google generate. Ads on which a user clicks remain private to Google and purchases that the same user makes at the advertiser’s store remain private to that company. We hope that others will start adopting this technology as well,” she added.

The Google Faculty Research Awards program aims to recognize and support world-class, permanent faculty pursuing cutting-edge research in areas of mutual interest. The goal is to identify and strengthen long-term collaborative relationships with faculty working on problems that will impact how future generations use technology.

According to Google, the Faculty Research Award is highly competitive - only 15% of applicants receive funding - and each proposal goes through a rigorous Google-wide review process. The award is structured as seed funding to support one graduate student for one year and is awarded as an unrestricted gift.

“The Google award is a recognition that Marina is doing advanced and relevant research that will enable many emerging AI/ML powered applications such as personal assistants and medical diagnosis where a delicate balance between effectiveness through customization and privacy need to be achieved,” said Chunming Qiao, professor and chair of the Department of Computer Science and Engineering.

Blanton’s research interests focus on information security, privacy and applied cryptography. In particular, her work spans across areas such as privacy-preserving computation and outsourcing, integrity of outsourced computation and storage, private biometric and genomic computation, authentication, anonymity and key management.

She joined UB’s Department of Computer Science and Engineering in 2017. Prior to joining UB, Blanton was an assistant professor in the Department of Computer Science and Engineering at the University of Notre Dame and a research scientist at the Air Force Research Laboratory in Rome, New York. She is a senior member of IEEE and ACM.

Blanton received her PhD from Purdue University in 2007.