The addRole-EA: A New Evolutionary Algorithm for the Role Mining Problem

Simon Anderer, Daniel Kreppein, Bernd Scheuermann, Sanaz Mostaghim

Abstract

Role Based Access Control is an important feature of cyber security and authorization management. The search of an optimum set of roles and their assignment to users can be modeled as an optimization problem, which is known as the Role Mining Problem (RMP) and has been shown to be NP-complete. Thus, fast heuristics are needed to search for solutions for the RMP in affordable time. This paper proposes a new evolutionary algorithm for the RMP, which is based on iteratively adding and deleting roles, while avoiding deviations from the original user-permission assignment. A range of experiments are presented which indicate that the proposed EA performs well on established benchmarking problems.

Download


Paper Citation