Using abductive and inductive inference to generate policy explanations

摘要

Providing reliable explanations for the causes of an access response represents an important improvement of applications usability and effectiveness, in a context where users are permitted or denied access to resources. I present an approach composed by two different procedures, both relying on OWL-DL and SWRL Rules, in order to generate policy explanations. The first procedure makes use of OWL Explanation and abductive reasoning. The second uses an algorithm of Association Rule Learning to identifying attributes and states arising together with policy privileges, in an inductive way. The PosSecCo IT Policy language is used in the present paper for representing the policies, but the approach is general enough to be applied in other environments as well.