Natural Sciences and Engineering Research Council of Canada
Symbol of the Government of Canada

Common menu bar links

Institutional links

Twitter LinkedIn YouTube Facebook

Chairholder Profile

Florian Kerschbaum

Florian Kerschbaum

David R. Cheriton School of Computer Science
University of Waterloo

Chair title

NSERC/RBC Industrial Research Chair in Data Security

Chair program

Industrial Research Chairs program

Role

Associate Chairholder since 2019

Summary

Canada’s future is dependent on its transition to a digital economy that is driven by data. Machine learning and data science enable data-driven decision-making to improve products and services and reduce environmental impact. However, the collected and generated data needs to be protected against threats to its confidentiality and integrity. Personally identifiable data that is protected by privacy legislation may be used to learn or make inferences, and the process may be further distributed over multiple parties under varying trust assumptions. The Associate Chairholder’s team, in collaboration with the sponsor, investigates data security algorithms and protocols in the entire life cycle of the machine learning and data science process, which enable the secure and safe use of the data in the digital economy while limiting the potential threats of its abuse.

The research will cover the protection of data during the learning and inference phase, consider all data, including training, test, model and prediction data, and comprehensively cover threats to confidentiality and integrity. Special emphasis is given to new threats to data that are emerging due to the rise of machine learning. The following are some examples. Machine learning models may be easily replicated. The research in this proposal will investigate the robustness of watermarking algorithms to track the provenance of the data and develop improved watermarking algorithms that can withstand the identified attacks. Reports about data breaches have become the daily norm, and criminals may use machine learning to exploit the stolen data. The proposed research will develop methods that protect stolen data against unwanted inferences. All of the proposed research will consider the conflicting objectives of security and privacy, efficiency, as well as utility and accuracy. On the one hand, the Industrial Research Chair (IRC) aims for practical algorithms and protocols consolidating existing research into novel applications useful to the sponsor and Canada. On the other hand, the IRC’s resources will be used to aim for breakthrough technologies that can significantly enhance the capabilities of the data scientist and have a significant impact on future scientific research.

The IRC is motivated by the University of Waterloo’s and Royal Bank of Canada’s (RBC) commitment to cybersecurity and privacy. Both organizations have formed a long-lasting partnership supporting the university’s Cybersecurity and Privacy Institute (CPI). The Associate Chairholder has been an associate professor in the David R. Cheriton School of Computer Science since 2017 and the CPI’s inaugural director. He has an outstanding international academic profile in data security and privacy and has extensive industry experience. The research will benefit RBC in its transformation to a data-driven company while upholding the highest standards of data protection. The Associate Chairholder will train a number of highly qualified personnel, potential future leaders in data security, closing the cybersecurity skills gap. The IRC is also designed to lift Canada’s entire financial sector to a higher security level in its transition to a digital economy with the help of the sponsor and the CPI.


Partner

  • Royal Bank of Canada

Contact information

David R. Cheriton School of Computer Science
University of Waterloo

Email: florian.kerschbaum@uwaterloo.ca

Website:
This link will take you to another Web site https://cs.uwaterloo.ca/~fkerschb/