Experimenting and improving perception of 3D rotation-based transitions between 2D visualizations

Abstract : Exploring a multidimensional dataset with visualization requires to transition between points of view. In order to enable users to understand transitions, visualization can employ progressive 3D rotations. However, existing implementations of progressive 3D rotation exhibit some perception problems with visualization of cluttered scene. In this paper, we present a first experiment showing how existing 3D rotation is effective for tracking marks, and that cluttered scenes actually hinder perception of rotation. Then, we propose to set the axis of rotation on the graphical marks of interest, and ran a second experiment showing that focus-centered rotation improves perception of relative arrangement.