b3.js: A Library for Interactive Web Data Visualizations in Virtual Reality

As virtual reality (VR) devices become more ubiquitous, data scientists increasingly recognize the power of viewing and manipulating complex, multivariate data in immersive virtual worlds. In their quest to extract knowledge from the data, data scientists may benefit from visualizations that are easily controlled by hand motion. We here present b3.js, a library for creating interactive VR data visualizations on a web browser. The 3D visualizations, for example, 3D scatter, bar, and surface plots, may be viewed through the Oculus Rift and manipulated gesturally with the Leap Motion controller. Oculus Rift provides immersive point of view perspective within the VR environment. Leap Motion enables gestural interaction by allowing users to control the parameters of the virtual camera, for example, modify the point of view. Users can also modify points of a rendered plot. We show that combining the Oculus Rift and Leap Motion with b3.js enables users to engage with a complex data set consisting of coherently-organized 3D flight paths of thousands of bats. A novel data exploration experience was created that is difficult to achieve on a flat screen with the traditional mouse-keyboard paradigm.