Rapid Evaluation of Techniques for Innovative Network Analysis (RETINA)

Intelligence analysts"ability to visualize, understand, and reason about terrorist organizations and other covert groups is limited by the current state of the art in network analysis techniques and tools. These techniques and tools work with homogeneous networks that consist of single types of both entity and relationship. They fail to work on heterogeneous networks that consist of multiple types of entities and relationships, details about both, and qualifying informatione.g., certainty, recency. Existing tools focus on presentation of information for use in briefings, rather than analytic support that enables analysts to develop new insights; or, they provide graph-theoretic methods, most of which are only useful in well-formed homogeneous networks. Qualifying information needs to factor into analyses and visualizations to engender analyst trust, otherwise analysts may misuse or ignore results. We propose to demonstrate a tool for Rapid Evaluation of Techniques for Innovative Network Analysis (RETINA), consisting of these elements: (1) a requirements analysis to understand the needs of analysts and the data they work with; (2) design of a network analysis framework to manage heterogeneous networks; (3) design, demonstration, and evaluation of innovative computational and visual analytic techniques; (4) design of techniques to map analytic results to actionable products.