Augmented Intelligence with Interactive Voronoi Treemap for Scalable Grouping: a Usage Scenario with Wearable Data

Ala Abuthawabeh, Abdelkader Baggag, Michaël Aupetit

Abstract: Interactive Voronoi Treemaps have been proposed to support arrangement and grouping tasks of data with snippet image representations. They rely on time-consuming manual actions to group data and cannot display more than a hundred images without occlusion. We propose visualizations designed to manage images visibility, evaluate group homogeneity, and shorten grouping task completion time while keeping control. It is supported by an automatic classifier forming an augmented intelligence system to tackle arrangement and grouping tasks at scale. We propose the usage scenario of a clinician using Interactive Voronoi Treemaps to group wearable data based on sleep visual patterns.