Securing XML with Role-Based Access Control

Abstract

Today's applications are often constructed by bringing together functionality from multiple systems that utilize varied technologies (e.g. application programming interfaces, Web services, cloud computing, data mining) and alternative standards (e.g. XML, RDF, OWL, JSON, etc.) for communication. Most such applications achieve interoperability via the eXtensible Markup Language (XML), the de facto document standard for information exchange in domains such as library repositories, collaborative software development, health informatics, etc. The use of a common data format facilitates exchange and interoperability across heterogeneous systems, but challenges in the aspect of security arise (e.g. sharing policies, ownership, permissions, etc.). In such situations, one key security challenge is to integrate the local security (existing systems) into a global solution for the application being constructed and deployed. In this chapter, the authors present a Role-Based Access Control (RBAC) security framework for XML, which utilizes extensions to the Unified Modeling Language (UML) to generate eXtensible Access Control Markup Language (XACML) policies that target XML schemas and instances for any application, and provides both the separation and reconciliation of local and global security policies across systems. To demonstrate the framework, they provide a case study in health care, using the XML standards Health Level Seven's (HL7) Clinical Document Architecture (CDA) and the Continuity of Care Record (CCR). These standards are utilized for the transportation of private and identifiable information between stakeholders (e.g. a hospital with an electronic health record, a clinic's electronic health record, a pharmacy system, etc.), requiring not only a high level of security but also compliance to legal entities. For this reason, it is not only necessary to secure private information, but for its application to be flexible enough so that updating security policies that affect millions of documents does not incur a large monetary or computational cost; such privacy could similarly involve large banks and credit card companies that have similar information to protect to deter identity theft. The authors demonstrate the security framework with two in-house developed applications: a mobile medication management application and a medication reconciliation application. They also detail future trends that present even more challenges in providing security at global and local levels for platforms such as Microsoft HealthVault, Harvard SMART, Open mHealth, and open electronic health record systems. These platforms utilize XML, equivalent information exchange document standards (e.g., JSON), or semantically augmented structures (e.g., RDF and OWL). Even though the primary use of these platforms is in healthcare, they present a clear picture of how diverse the information exchange process can be. As a result, they represent challenges that are domain independent, thus becoming concrete examples of future trends and issues that require a robust approach towards security.