Role Mining in Distributed Firewall Using Matrix Factorization Methods

Role mining is the process of extracting roles in an organization from available input. Firewall rules are focused and taken as input and extract roles from the same. Role mining helps in creating a Role-Based Access Control (RBAC) system from a discretionary access control (DAC) system. The firewall rule will be represented in a three-dimensional matrix and from the roles are extracted. The above process is implemented over a distributed firewall. User Permission assignment (UPA) matrix will be used for the reduction process. The central machine in the distributed architecture will push firewall rule to the intended destination. Policy files can have a hierarchical firewall rule of different machines. A comparison is done on various matrix factorization algorithm available, From the same, a faster one is being selected and used for extraction of roles. The faster algorithm can be used in an environment where firewall rules are updated frequently. A better matrix factorization algorithm is aimed at extracting a role in an organization. The mining role from firewall rule is having applications in all the organizations having networks implemented based on the roles of people. The firewall rule gives an insight into the structure of an organization.