Dynamic document reclassification for preventing insider abuse

Digital documents in an organization are usually classified into static secrecy levels such as top-secret, secret, confidential and unclassified. Factors such as changes in the user hierarchy and addition of new projects generally require a change in a document's importance. Enforcing such changes in relative importance (RI) of documents protect the privileged documents from insider abuse. In this paper we propose a new framework for monitoring and dynamically changing the RI of documents over time. The proposed scheme utilizes a reasoning technique which takes into consideration the organization-specific data and the document usage data to make an informed decision on the RI of documents. Various inputs to the reasoning framework are standardized using XML schemas to provide interoperability. We have performed simulations with synthetic document usage data and report the results as proof of concept.