Our research on data privacy is centered on designing solutions to  enable privacy preserving data analytics. We study different representations of data drawn from a variety of computing application scenarios that include: small datasets drawn from resource-constrained scenarios, high-dimensional data, unstructured data, and image data. From this basis, our work encompasses research on sensitive information discovery in data, quantifying privacy exposure risks, usability, and the broader challenges spanning the characterisation of privacy subversion exploits as well as the impact/effectiveness of distortions in guarding against privacy subversions.

On-going projects include works such as (1) Quasi-Identifier Discovery, (2) Outlier/Anomaly Detection, and (3) Automated PII Discovery in Composed Data (AutoPII).