CBARMM: Cluster-Based Association Rule Mining Method for Attribute-Based Access Control

Attribute-Based Access Control (ABAC) technology utilizes the inherent attributes of subjects, environments, and objects for effective access control. It is characterized by dynamic flexibility, strong scalability, granularity, and automated decision-making, making it an ideal solution for data security in big data environments. However, transitioning from traditional access control models, such as Access Control Lists (ACL) and Role-Based Access Control (RBAC), to ABAC can be a time-consuming, labor-intensive, and error-prone process, particularly when configuring security policies. Based on this, we propose the cluster-based association rule mining method (CBARMM) to automatically generate security policies from access logs and attribute files, thereby simplifying the model migration process. First, we employ the random forest algorithm to select attributes and evaluate their impact on authorization decisions. Next, we apply weight based K-modes and SD-map algorithms to extract attribute relationship filters and attribute value filters. Finally, the policy is continuously optimized through two methods of policy refinement. The experimental results demonstrate that our proposed method holds significant research and application value in the mining of attribute-based access control policies.