Dynamically Distributing Diagrams Across Devices
On visualizations that organize themselves
Today’s digital workspaces are evolving rapidly, and so are the ways people interact with data. Consider the daily workflow of an oncologist. In the morning, she reviews patient data on her smartphone during her commute. At the office, she uses a desktop and a tablet to plan treatments. Later, she discusses cases with colleagues using large displays and laptops at a tumor board meeting. Finally, she consults with patients in a treatment room, using a tablet and a large TV. The only constant thing with this oncologist’s workspace is that it is never constant.
In 2019, at the CHI conference in Glasgow, Scotland, my colleagues and I presented Vistribute, a framework for automatic visualization layout in this kind of multi-device environment. Using Vistribute, our oncologist friend can seamlessly continue her data exploration across all her devices without having to manually set up the interface every time.
At its core, Vistribute is about making data visualization more flexible and accessible. Our framework consists of three main parts: a design space that considers the properties and relationships of interactive visualizations, devices, and user preferences in multi-display environments; six specific heuristics to guide the distribution of visualizations across different devices; and a web-based implementation that automatically generates a suitable layout distribution for a given interface and device ensemble, allowing users to make further adaptations if necessary.
Multi-device environments are becoming increasingly common, but are woefully underused. If a new device is at all used, it is only to increase screen real estate. However, Vistribute attempts to harness the potential of multiple devices by dynamically adapting to different devices, settings, and collaborators. The automatic approach means that analysts can continue their work uninterrupted, regardless of the devices they switch between and the collaborators they meet throughout the day.
What makes Vistribute unique is its ability to infer all necessary information by analyzing the visualizations and devices involved without relying on additional input from users or programmers. This not only saves time but also enhances the quality and consistency of the visualizations.
To demonstrate the validity of the framework, we conducted a qualitative study where we allowed experts to create their own visualization layouts and then rate both manual and automatic ones. The results were promising — Vistribute’s automatic distributions were comparable in quality to those created manually by experts. This underscores the framework’s effectiveness and potential for widespread adoption.
Since its introduction at CHI 2019, Vistribute’s relevance has grown with the increasing ubiquity of multi-device environments in both professional and personal settings. With more than 50 citations (Google Scholar, June 2024), the principles and technology behind Vistribute continue to inspire new applications and developments in the field of data visualization and human-computer interaction.
For those interested in diving deeper into this research, we encourage you to explore our full paper:
- Horak, T., Mathisen, A., Klokmose, C. N., Dachselt, R., & Elmqvist, N. (2019). Vistribute: Distributing Interactive Visualizations in Dynamic Multi-Device Setups. In Proceedings of the ACM SIGCHI Conference on Human Factors in Computing Systems. ACM, New York, NY, USA. https://doi.org/10.1145/3290605.3300846
