![](https://d3ilqtpdwi981i.cloudfront.net/aQ5mGffAj87v72Q0X4dsoBw41fA=/425x550/smart/https://bepress-attached-resources.s3.amazonaws.com/uploads/be/f7/08/bef70810-6cbd-482a-a7f9-e7f4de6e7981/thumbnail_a30fb268-4bcc-4cc8-896a-3effa1104b7b.jpg)
Verification of legitimate access of documents, which is one aspect of the umbrella of problems in the Insider Threat category, is a challenging problem. This paper describes the research and prototyping of a system that takes an ontological approach, and is primarily targeted for use by theintelligence community. Our approach utilizes the notion of semantic associations and their discovery among a collection of heterogeneous documents. We highlight our contributions in (graphically) capturing the scope of the investigation assignment of an intelligence analyst by referring to classes and relationships of an ontology; in computing a measure of the relevance of documents accessed by an analyst with respect to his/her assignment; and by describing the components of our system that have provided early yet promising results, and which will be further evaluated more extensively based on domain experts and sponsor inputs.
Available at: http://works.bepress.com/amit_sheth/49/
Presented at the IEEE International Conference on Intelligence and Security Informatics, Atlanta, GA, May 19-20, 2005.
Attached is the unpublished, author's version of this proceeding. The final publication is available at Springer via http://dx.doi.org/10.1007/11427995_47.