Leveraging Machine Learning to Improve Access Control Mechanisms in Data Warehousing

Abstract

This paper delves into integrating Machine Learning (ML) algorithms into Access Control Mechanisms (ACM) within Data Warehouses (DW) to enhance both security and operational efficiency. Traditional ACMs, notably role-based access control (RBAC), often struggle with adapting to dynamic threats and complex access patterns. ML tackles these issues by improving access decision accuracy through real-time anomaly detection and adaptive control based on user behavior and contextual insights. The study examines the implementation and benefits of ML-enhanced ACMs in DW environments, highlighting their effectiveness in mitigating unauthorized access and insider threats. This research also addresses critical considerations, such as privacy protections and the interpretability of ML models, crucial for maintaining regulatory compliance and stakeholder trust. Ultimately, this study emphasizes ML's pivotal contribution to advancing security practices within Data Warehouses (DWs) and proposes future research directions to enhance adaptive security measures in evolving digital landscapes.