An evolutional model for operation-driven visualization design

• Published in: Journal of visualization 2022-02, Vol.25 (1), p.95-110
• Main Authors: Wei, Yating, Mei, Honghui, Huang, Wenqi, Wu, Xiangyang, Xu, Mingliang, Chen, Wei
• Format: Article
• Language: English
• Subjects: Classical and Continuum Physics; Composition; Computer Imaging; Design; Engineering; Engineering Fluid Dynamics; Engineering Thermodynamics; Flow charts; Heat and Mass Transfer; Pattern Recognition and Graphics; Regular Paper; Vision; Visualization
• ISSN: 1343-8875; 1875-8975
• DOI: 10.1007/s12650-021-00784-w

Although there are many visualization authoring tools, often the design process is confirmatory in that the user either selects and refines pre-built visualization templates or manually constructs the visualization following the standard visual mapping pipelines. During this process, an amount of time is spent on the interactive specification of visual forms or visual encoding configurations. In this paper, we present a novel visualization design scheme that provides operational guidance for progressive refinements when designing a visualization. The key idea is an evolutional operation model that models the design flowchart as a composition of a sequence of operations based on the current design and recommends available evolutions in the design space. One distinct advantage of our approach is that it facilitates discovering design subtasks relevant to the visualizations and enables the composition of a visualization based on the evolutional operations. As a demonstration of the concept, we implement a prototype and test its efficiency on a set of challenging design examples. Case studies verify that the proposed scheme effectively enhances the creativity of designs and releases the user from having to interpret their data and design intention with toolkit-specific abstractions. Graphic abstract