Visual analytics is a multidisciplinary field of science that involves analytical reasoning facilitated by interactive visual interfaces. Our research aim is to increase awareness of important visual analytics problems and improve evaluation methodologies for complex visual analytics systems. We accomplish this by creating real world scenarios with accompanying synthetic datasets. These datasets contain embedded ground truth and known solutions so that professionals can effectively evaluate new visual analytics tools and techniques.

We decided to leverage the Visual Analytics Science and Technology (VAST) challenge framework for the development of scenarios and synthetic datasets. We began by investigating problems specific to the bioinformatics domain. To help understand the data analysis process from the biological perspective, we first created a visual analytics model for lead generation library design in drug discovery. Combined with our experience from prior VAST challenges, we introduced a 2010 mini-challenge based on a hypothetical pandemic outbreak involving a rapidly evolving fictitious virus. We participated in the entire challenge process including dataset generation, challenge administration, and judging. After evaluating the results, we designed an improved scenario for the 2011 challenge. This iteration focused on epidemic outbreaks caused by dangerous biological agents. Finally, we present a series of recommendations for future VAST challenge administrators. With these recommendations, our aim is to enable the design of scenarios that push the forefront of visual analytics research relevant to real-world biomedical research problems.