ASPIRE is a research project that aims to develop novel privacy and data protection methods that leverage Artificial Intelligence synthetic data generation.
Advancements in data-driven models across various fields such as education (learning analytics and Artificial Intelligence in Education) and healthcare have been fuelled by a breadth of large datasets. However, the growing demand for large volumes of data comes with significant challenges, with privacy concerns being among the most critical.
Traditional methods such as anonymisation and de-identification have become increasingly vulnerable to attacks, which increases the risk of personal information exposure. In response, synthetic data is emerging as a promising and novel solution providing scientists and researchers with an alternative to real data without compromising privacy or exposing sensitive information.
Primary Objective
ASPIRE aims to develop novel synthetic data generation methods that robustly ensure privacy and data protection using two innovative techniques: Differential Privacy and Federated Learning. By integrating both new privacy-preserving technologies of differential privacy and federated learning, the project will accelerate the reliability and security of synthetic data generation.
By leveraging Artificial Intelligence and Synthetic data, the project aims to develop innovative methods for sharing and analysing data without compromising the privacy of the individual. ASPIRE will have an impact in two key areas of education and healthcare and aim to foster a more trustworthy environment for data exchange.
The project will hold potential for creating a trustworthy dataset exchange system, contributing to the future of privacy-preserving technologies.
ASPIRE cutting-edge approach has the potential to reshape how sensitive data is handled, and may offer a scalable solution that balances the urgent need for data privacy with the growing demand for data-driven insights.
PhD Position
The project includes a PhD position for a fixed term of 3 years, starting in Q1 2025. Synthetic data generation is a key driver for the PhD position.
The current ASPIRE project team consists of Project Leader Mohammad Khalil and Researchers Qinyi Liu and Farhad Vadiee. The project team and the PhD candidate are expected to offer novel perspectives on proposing rigorous privacy guarantees by integrating differential privacy and federated learning with synthetic data. Additionally, an evaluation of the utility of the synthetic data will be carried out, to ensure that it remains functional while adding an extra layer of data protection.
