SBIR-STTR Award

We propose a network visualization technique that uses multiple, meaningful dimensions, displayed in a series of projections that simplify the display while retaining latent information. This multidimensional system simplifies the visual display in ways that enhance comprehension, enable analysis of the evolution of the network over time, and facilitate representation of uncertainty. Throughout, the system provides both algorithmic identification of complex network structures, and interactive visualization that enables discovery by the analyst. We show that the multidimensional approach is more useful than the traditional approach of placing nodes in accordance with their relationships with other nodes. Our use of information-bearing dimensions provides the analyst with meaningful visual context, enabling the visual interpretation of network structures and enabling better understanding of navigation through the network. While relationally organized displays coalesce conflicting or changing information, our use of time as a dimension allows us to retain information about previous states of the network, while our use of node collections helps us analyze uncertain and conflicting information.

Keywords:
NETWORKS, VISUALIZATION, INTERACTIVE GRAPHICS, DYNAMIC GRAPHS, UNCERTAINTY, Q-ANALYSIS, MULTIDIMENSIONAL PROJECTIONS, TRUST METRICS