Optimized trusted information sharing

Abstract

Optimized Trusted Information Sharing Joseph M. D’Alessandro As the digital world expands the building of trust and the retention of privacy in information sharing becomes paramount. A major impediment to information sharing is a lack of trust between the parties, based on security and privacy concerns, as well as information asymmetry. Several technological solutions have been proposed to solve this problem, including our’s: a trusted enclave with a Continuous Compliance Assurance (CCA) mechanism. Of the work surrounding these proposed solutions, no attention has been directed toward studying the issues of performance surrounding processing of this nature. Studies have shown that ignoring the performance of a system can lead to ineffectiveness (i.e. disabling certain features), and can be severely detrimental to system adoption. To ensure that our trusted enclave and CCA mechanism are viable solutions to the trusted information sharing problem, we have built a prototype CCA mechanism and a test bed. The test bed has allowed us to identify problem areas within our prototype. One such area is compliance verification, which utilizes the XPath language in order to test XML encoded information for compliance to regulatory and corporate policies. The compliance verification problem can be described as the answering of multiple queries over a single XML document. We proposed and tested multiple state-of-the-art algorithmic as well as system-based improvements to XPath evaluation, in order to better the overall performance of this aspect of our system. We integrated each of the improvements into our prototype mechanism and have observed the results. Our experiments have taught us much about the problem of compliance verification, and has led us in new directions as we continue to search for a solution.