Adapting NIEM input for automated privacy policy reasoning

The government information sharing community faces many challenges facilitating compliant information sharing and usage across domains, including complex legislation that is difficult for non-specialists to interpret and verify by hand. We have demonstrated the utility of AIR (Accountability in RDF), an automated policy reasoning system, in the past to encode such rules in computer-readable format and reason over them using facts represented in RDF. However, RDF is not widely adopted as a serialization mechanism for the types of data that an AIR-based system would take as input. The RDFIFY conversion layer, which we constructed as part of this work, produces an RDF representation of data presented in NIEM (National Information Exchange Model) format, an XML-based serialization used in the government community. When incorporated into an AIR-based system, RDFIFY permits an unmodified AIR reasoner with appropriate rules to consume and reason meaningfully over data presented in this format, allowing the AIR technology to be applied to existing government information exchange environments. In performing this work, we also encountered a limitation of the NIEM format in comparison to native RDF, namely a lack of standardized references to objects defined in external sources, which is of particular relevance in the context of information sharing in a distributed environment, but can be readily resolved with minor adjustments to the NIEM standard.