The goal of this project is to develop technology in support of the visual analytics process. The complex process of analysis in domains from scientific discovery to homeland security consists of cycles of hypothesis generation, evidence gathering, evidence organization, and hypothesis resolution. The developed discovery management technology provides support, both visually and computationally, not only for all analysis stages in isolation, but also for tight integration among them. The computational infrastructure enables analysts to capture, visualize, manage, analyze, and interactively explore patterns and discoveries ("nuggets") of the visual analytics process. Tools for the modeling and management of nuggets and their complex interrelationships form the foundation of the infrastructure. Methods for nugget generation include explicit identification and confirmation by user, implicit capture based on analysis of user logs, and automated discovery using statistical and data mining techniques.
Computational and interactive visual methods enable analysts to efficiently validate, annotate, classify, organize, and purge nuggets over time. Visual representations of hypotheses, evidence, and nuggets help analysts explore their data, manage their discoveries, and organize their reasoning processes. The resulting technology enables analysts from diverse domains ranging from biomedical discovery to homeland security to visually explore data, to form hypotheses, and to swiftly manage their discoveries as the exploration process proceeds. Project insights are infused back into an educational agenda using a rich repertoire of mechanisms, ranging from special-topics courses to undergraduate and K-12 project-based student activities. The project Web site (http://davis.wpi.edu/~xmdv) is used for results dissemination.
