Strengthening National Digital Infrastructure Privacy Focused Data Pipelines for Ethical Behavioral Analytics

Abstract

In this paper, the researcher will analyze the performance of a privacy committed data pipeline against a standalone centralized one on large data behavioral analytics. The findings indicate that privacy controls introduce small overheads but performance of the system is not lost in a serious manner. The predictive models are also precise, and there is less than a 5 percent loss even in case of the different privacy, synthetic data, or federated analytics. High privacy improvements were realized, such as significant re-identification risk reduction and violation of policy. Telecom and commerce cross-sector testing process assures that the design is generic and is applicable to any dataset. The results indicate that high analytical performance and strong privacy may co-exist.