Project Website
Publications
Data Animator: Authoring Expressive Animated Data Graphics, CHI 2021
John Thompson, Zhicheng Liu, and John Stasko
Understanding the Design Space and Authoring Paradigms for Animated Data Graphics, EuroVis 2020
John Thompson, Zhicheng Liu, Wilmot Li, and John Stasko
Overview
Our goal is to create next-generation authoring tools for expressive animated data graphics. Creating such graphics often requires designers to possess highly specialized programming skills. Alternatively, the use of direct manipulation tools such as Adobe After Effects is popular among animation designers, but these tools have limited support for generating graphics driven by data.
To understand the design space of animated data graphics, we survey real-world examples and characterize animated transitions based on object, graphic, data, and timing dimensions. We synthesize the primitives from the object, graphic, and data dimensions as a set of 10 transition types, and describe how timing primitives compose broader pacing techniques. We then conduct an ideation study that uncovers how people approach animation creation with three authoring paradigms: keyframe animation, procedural animation, and presets & templates. Our analysis shows that designers have an overall preference for keyframe animation. However, we find evidence that an authoring tool should combine these three paradigms as designers’ preferences depend on the characteristics of the animated transition design and the authoring task.
Based on these findings, we present Data Animator, a system for authoring animated data graphics without programming. Data Animator leverages the Data Illustrator framework to analyze and match objects between two static visualizations, and generates automated transitions by default. Designers have the flexibility to interpret and adjust the matching results through a visual interface. Data Animator also supports the division of a complex animation into stages through hierarchical keyframes, and uses data attributes to stagger the start time and vary the speed of animating objects through a novel timeline interface. We validate Data Animator’s expressiveness via a gallery of examples, and evaluate its usability in a re-creation study with designers.