An Efficient Solution to User Authorization Query Problem in RBAC Systems Using Hierarchical Clustering

 Role Based Access Control (RBAC) systems face an essential issue related to systematic handling of users’ access requests known as the User Authentication Query (UAQ) Problem. In this paper, we show that the UAQ problem can be resolved using Unsupervised machine learning following the guaranteed access request and Dynamic Separation of Duty relations. The use of Agglomerative Hierarchical Clustering not only improves efficiency but also avoids disordered merging of existing roles to create new ones and steers clear of duplication. With a time complexity of  O(n^3), the algorithm proves to be one of the fastest and promising models in state-of-the-art. The proposed model has been compared with the existing models and experimentally evaluated.