Mike Rosulek
Mike Rosulek, assistant professor of computer science

Mike Rosulek, assistant professor of computer science in the College of Engineering at Oregon State University, received a Visa Faculty Research Award to advance methods of customer privacy and fraud detection.

“We’ve known for several decades that cryptography can protect not just data at rest, but also data in use, at least in principle. Finally, in the last several years these cryptographic ideas have been improved to become truly practical,” Rosulek said.

Rosulek and his colleague at Visa, Payman Mohassel, will be working to improve a tool from cryptology called private set intersection, which allows two parties to find items in common on two separate lists without revealing any other information from the lists.

Their research will help make complicated queries faster to process. For example, a company may want to know how many customers they have in common with another company without revealing who those customers are.

The funds will support one graduate student for a year who will be helping to develop new prototypes that would make advanced cryptography practical for companies.

“This award demonstrates that industry leaders see the potential of advanced cryptography to protect data during use and solve real-world privacy challenges,” Rosulek said.

Mike Rosulek

Mike Rosulek, assistant professor of computer science, was selected for a Google Research Award for a grant to advance methods of customer privacy. The award will allow Rosulek to hire a graduate student to work on the project, and give them the opportunity to collaborate with Google researchers and engineers.

“I have lots of ideas in this space, and it will allow me and a student to dive in head first exploring them,” Rosulek says.

Companies are looking for inexpensive ways to share information with each other without violating the privacy of their customers. For example, two companies may want to find out which customers they have in common. A tool from cryptology called private set intersection allows two parties to find items in common on two separate lists without revealing anything else from those lists.

One part of Rosulek’s research seeks to strengthen the security of private set intersection tools while keeping the costs reasonable so that companies are more likely to adopt good practices for keeping their customer’s information secure.

Another part of the project will work on flexible (or “fuzzy”) matching of items on lists such as addresses. Names and street addresses may have differences in spelling, so looking for exact matches between two sets can be too restrictive. Rosulek’s research will seek to modify current techniques to allow for “close enough” matches.

“I’m excited to see that Google is interested in these advanced cryptographic tools. I’m excited that the techniques can be used to protect sensitive user information. And I’m excited about the new technical and mathematical challenges on the roadmap,” Rosulek says.