An improved minimal noise role mining algorithm based on role interpretability

Abstract

Interpretable role mining has achieved notable improvements in usability and effectiveness of roles in RBAC deployments, owing to its virtue in mining meaningful roles. However, current research ignores the interference caused by data noise, thus limiting broader application and deployment in real scenarios. In this paper, the interpretable role mining problem is extended considering data noise, referencing the minimal noise role mining problem. Accordingly, an improved minimal noise role mining algorithm is proposed to optimize the reconstruction error and role interpretability. The experimental results on real data demonstrate that the proposed algorithm has better efficiency, lower reconstruction error while ensuring the interpretability regardless of the data scale.