Keywords: visual analytics, data science, multivariate data visualization, temporal data visualization, machine learning
Due to the exploratory nature of scientific studies involving data from Department of Energy user facilities and leadership computing centers, scientists still require human-in-the-loop tools that allow them to ask questions and generate new hypotheses beyond the ideas that preceded the data collection. However, the size and complexity of scientific data preclude a completely manual and exhaustive approach. Therefore, scientists require both automated computational guidance to reduce the vast search space and flexible human interaction techniques to efficiently explore their data. In this talk, I will describe two interactive data visualization and analysis systems that I have developed in collaboration with experts from the fields of additive manufacturing and climate science. In addition to describing the features of these systems, I will highlight new discoveries that this research has enabled for challenging domain-specific problems. I will also describe the interdisciplinary, participatory design process that I employ in my projects. I will conclude with an overview of my future research objectives, which I hope will further enhance the abilities of humans to explore, understand, and communicate insight from today’s large and complex data.