Please note: This PhD seminar will be given online.
Karl Knopf, PhD candidate
David R. Cheriton School of Computer Science
Supervisor: Professor Xi He
Differentially private query answering systems have been designed to support interactive exploration of sensitive data. When using differential privacy, the privacy budget is the only parameter that can control the trade-off between the loss of sensitive information and utility of the query answers. Under a limited privacy budget it is possible that not all queries will be able to be answered as accurately as desired by the data analyst.
Prior work has proposed to relax the privacy guarantees by using policies to specify a smaller set of sensitive information. The policy-aware approach allows for a better utility-privacy trade-off than standard differential privacy. A policy-aware differentially private algorithms need to represent the policy as a graph over the full domain of the datasets. This is often inefficient in an interactive setting where the datasets frequently have high dimensions.
To join this PhD seminar on Zoom, please go to https://us02web.zoom.us/j/83326411204?pwd=Z3dNVUxIK01PMXY3MTlXaHNVckJqdz09.
200 University Avenue West
Waterloo, ON N2L 3G1