This paper presents a judgement and decision making analysis of collaborative problem solving. This analysis is done with respect to the Raven and CoRaven decision-making tools for filtering, interpreting, and visualizing large amounts of uncertain data in the domain of intelligence analysis. Raven and CoRaven are multimodal advisory decision aids and their inferential reasoning is based on Bayesian networks. Human decision makers and information sources interact with these decision making systems in many ways, during their design, construction, refinement, and usage. This paper analyzes the collaborative aspects of the use of Raven and CoRaven using the Judge Advisor System model.
David C. Wilkins, Patricia M. Jones, Roger Bargar,