Enhancing Data Governance through Data-Centric AI: Case Study in New Zealand Government Sector

Alex X. Wang, Colin R. Simpson, Binh P. Nguyen

Abstract: Data governance is crucial in the government sector, emphasising data privacy and ethics. The rise of synthetic data in Data-Centric AI has the potential to address data quantity, quality, and availability concerns. However, its application in enhancing data governance within the government sector remains overlooked. To fill this gap, we propose a comprehensive frame-work for utilising synthetic data to enhance data governance. The proposed framework comprises three key components: real data profiling, synthetic data generation, and evaluation. Real data profiling identifies potential confidentiality risks in the data and offers statistical guidance for subsequent generation steps. The synthetic data generation and evaluation component ensures that privacy-sensitive synthetic data is generated when real data is unavailable. To demonstrate the effectiveness of our framework, we conducted experiments in the New Zealand government sector. The results clearly indicate the robustness of the framework. Consequently, we conclude that synthetic data can serve as a valuable addition to the existing toolkit for enhancing data governance in the government sector. By leveraging synthetic data, the government can better address data challenges and uphold data privacy and ethical standards more effectively.