Authorization Robot 

From vision to the product

What started as a research project in 2017 is now a unique product. In collaboration with the German Federal Ministry of Education and Research and Karlsruhe University of Applied Sciences (HKA), we developed the first virtual ERP roll builder.

• The challenge

  When implementing the automatic authorization concept, the challenge is that the new software must be able to process very large amounts of data. These data volumes are too large for the standard data analysis procedures available in the scientific literature. Efficient matrix clustering methods are used here, which identify related departments/processes/patterns in the usage data as preprocessing, so that this enables partitioning with as little information loss as possible and reduces the problem size.

• The evolutionary algorithm

  On the basis of this partitioned usage data, an evolutionary algorithm is used to generate the best possible role concept. This "role mining problem", which is well known from the literature, is assigned to the complexity class NP-complete. Therefore, the evolutionary approach as a metaheuristic method provides the right tool to approach an optimal solution efficiently and continuously.

• The evaluation criteria

  In order to provide the evolutionary algorithm with suitable evaluation criteria for the role concepts, it requires additional complex modules that work with the main algorithm. Compliance can be mentioned here as an example. A set of rules for all audit-relevant conflicts and critical authorization combinations that allows the compliance of a role concept to be evaluated and optimized accordingly.

• The semantic meaningfulness

  A much more difficult undertaking is the automated evaluation of the semantic meaningfulness of the roles. In other words, are the roles that have been generated also nameable, i.e. could a human give the role and its content a semantically meaningful name? Here, we use a machine learning method, more precisely neural networks, to learn from past projects and other sources the experience of consultants and experts, so that this module can evaluate whether a role can be named by a human and if so, which name makes sense.