Database roles analysis using data mining

Role based access control (RBAC) has been around for several decades now. Role design has a very strong impact on database security and it could be the source of many security incidents by accidental or intentional malicious activity of authorized users. Data mining has been used for knowledge discovery in databases and data warehouses. It is efficient for discovering patterns and extracting statistically important metadata. This paper proposes a method for database roles analysis using data mining. Implementing Frequent Pattern Growth algorithm (FP-growth), divide and conquer approach is applied and roles' frequent subsets are determined. Using these subsets, further analysis is much simpler and results in higher security levels for accessing sensitive data.