Towards Automated Access Control Policy Mining via Structured Attribute-Based Access Control

Creating and maintaining access control policies is a challenging problem for security administrators. To address this problem, we present an improved algorithm to mine access control policies from potentially incomplete logs, via structured attribute-based groupings of principals and resources. Preliminary experimental results using a proof-of-concept implementation show that the miner provides a foundation for the construction of structured dynamic policies. The objective of this system is to respond to changes impacting access control decisions on the fly and to strengthen security, facilitate ease of use and provide greater flexibility when dealing with complex organisational networks.