The RBAC System Based on Role Risk and User Trust

Abstract

Abstract:The risk problem of access control model is always the key point of security management. In this paper, we introduce the risk factor to control the risk of role under the proposed threshold when building RBAC policy based on role mining. At the same time, a user should meet some demands to activate roles. So we propose the following principal. A user could activate a role when his trust is higher than the trust threshold of the candidate role. This method makes role authority more reasonable and it avoids of the abuse of permissions. Experiments on performance indicate that the proposed algorithm can not only build the complete RBAC system to avoid roles' high risk caused by the high complexity of permissions in roles, but also meet the least privilege when a user activates a role.

Optimization of Access Control Policies

“…The risk of an ACP reflects the aggregated risk of the permissions that are permitted by it. ACP modelling or optimization methods can use risk minimization as an objective to reduce the impact of excessive permission assignments (Jin et al, 2016;Dos Santos et al, 2014). High risk is also an indicator for high maintenance priority and can serve as context information for policy engineers and reviewers, based on the assumption that a high risk value suggests a more restrictive handling than a low one (Fuchs et al, 2014).…”

A Conceptual Model of Role Based Access Control Using Role Mining Algorithm

“…The bottom-up approach, according to Hu et al [8] can uncover roles from the existing user-permission assignments (UPA) automatically that is known as role mining and because of its nature that based on computing-intensive approach, it is widely applied to build a RBAC model. However, to build and sustain a RBAC model, role mining is becoming a great interest [9][10] and the authors have identified the need of role mining to design and develop an algorithm to determine roles based on data mining methods because it could reduce the cost of allocating roles manually thus able to construct a concise RBAC system. The next section would provide in-depth analysis on methodology to build a RBAC model using role mining algorithm.…”

“…e) RH ⊆ R × R, a partial order on roles described the inheritance relationships. RBAC state could be expressed as <R, UA, PA, RH> that is consistent with an access control configuration =< , , >, where U defines a set of all users, P is a set of all permissions and ⊆ × as the user-permission relation [10].…”