Dr. Reza Samavi
Areas of Academic Interest
Information security
Machine learning
Trustworthy AI
Education
| Year | University | Degree |
|---|---|---|
| 2013 | University of Toronto | PhD |
| 2007 | University of Toronto | MEng |
| 1987 | Amirkabir University of Technology | BSc |
Spotlight
During the dot com boom, Reza Samavi founded his own successful start-up in software development. Then after a decade, he started rethinking what kind of impact he wanted to make in his work and switched to academia. “The pace of change in software was so quick and fundamentally impacting the way we socialize and make decisions,” says Samavi. “I began thinking about the ethical and social implications and realized I couldn’t develop new products without fully understanding their impact.”
Now as a professor, Samavi’s research lies at the intersections of artificial intelligence and machine learning, security and privacy. While most AI research is focused on the performance of AI, Samavi’s work examines if the algorithms are trustworthy, ethical and fair. “In the future, for example, rather than a committee hiring a new employee, an AI algorithm could decide,” says Samavi. “It could be very dangerous if we don’t trust that solution. In the next 10, 20 and 30 years, these will be the fundamental questions as we incorporate AI into society.”
“Beyond teaching and research, what I really enjoy about academia is forming a real human connection to students.”
- Privacy By Design Ambassador, designated by the Information and Privacy Commissioner of Ontario
- Privacy Technology Research Award, IBM Center for Advanced Studies (CAS)
- Privacy by Design Research Award, Information and Privacy Commissioner of Ontario
- Sutton, Andrew, and Reza Samavi. "Tamper-Proof Privacy Auditing for Artificial Intelligence Systems." IJCAI, pp. 5374-5378. 2018.
- Liang, Yuting, and Reza Samavi. "Optimization-based k-anonymity algorithms." Computers & Security 93 (2020): 101753.
- Samavi, Reza, and Mariano P. Consens. "Publishing privacy logs to facilitate transparency and accountability." Journal of Web Semantics 50 (2018): 1-20.
- Sutton, Andrew, and Reza Samavi. "Blockchain enabled privacy audit logs." International Semantic Web Conference, pp. 645-660. Springer, Cham, 2017
- Sutton, Andrew, Reza Samavi, Thomas E. Doyle, and David Koff. "Method for enabling trust in collaborative research." U.S. Patent Application 16/552,705, filed March 5, 2020.
- Member, Professional Engineers Ontario
- Member, IEEE
- Member, Association for Computing Machinery (ACM)
- Member, International Association of Privacy Professionals (IAPP)
- Using game theory to model poisoning attack scenarios, Interview with TechXplore - Machine Learning & AI. Article, June 2019. (external link)
- Privacy investigations. Creating a model to audit data protection practices, Interview with the science writer of Computer Science journals at Elsevier. Article, November 2018.