StructGraphics: Flexible Visualization Design through Data-Agnostic and Reusable Graphical Structures

Abstract

Fig. 1. The StructGrapher's user interface consists of: (a) a library of reusable visualization structures, (b) a visualization sketcher for drawing basic shapes and grouping them into collections, (c) a property inspector for exposing and structuring graphical properties, and (d) a spreadsheet for creating mappings between properties and data. The approach enables designers to construct data-agnostic visualizations and create flexible bindings with data. The bottom visualization design was originally published in the Financial Times [22].

Generative Design Inspiration for Glyphs with Diatoms

“…Moreover, these tools could benefit from the addition of interactive design externalization and of small permutables in particular: the ability to see the same data point visualized many different ways across a single display. For example, this externalization could manifest as a peripheral or collapsible view akin to StructGraphics' gallery of saved templates [79]. This automated externalization could reduce the need to manually capture and arrange design artifacts in external tools.…”

“…Recent interactive visualization construction tools allow people to craft bespoke visualization beyond conventional statistical charts. These tools include Lyra [63], Data Illustrator [46], Charticulator [58], and most recently StructGraphics [79]. Whether the goal is to realize mash-ups of existing chart types or to draw xenographics ("weird but (sometimes) useful charts" [44]), the output of these authoring tools typically serves a specific communicative intent [41], where transferability of the output to other datasets is not as critical as novelty and memorability [11].…”

Generative Design Inspiration for Glyphs with Diatoms

“…Moreover, these tools could benefit from the addition of interactive design externalization and of small permutables in particular: the ability to see the same data point visualized many different ways across a single display. For example, this externalization could manifest as a peripheral or collapsible view akin to StructGraphics' gallery of saved templates [80]. This automated externalization could reduce the need to manually capture and arrange design artifacts in external tools.…”

“…Recent interactive visualization construction tools allow people to craft bespoke visualization beyond conventional statistical charts. These tools include Lyra [64], Data Illustrator [47], Charticulator [59], and most recently StructGraphics [80]. Whether the goal is to realize mash-ups of existing chart types or to draw xenographics ("weird but (sometimes) useful charts" [45]), the output of these authoring tools typically serves a specific communicative intent [42], where transferability of the output to other datasets is not as critical as novelty and memorability [12].…”

No Grammar to Rule Them All: A Survey of JSON-style DSLs for Visualization

“…Similarly, only a small set of mostly industry-led DSLs provide their own accessibility features (Highcharts, Vega-Lite, ECharts, and FusionCharts). This dearth may be due to the fact that such practicalities are often viewed as mere implementation details rather than being central to usability [82,107], which may be compounded by a lack of research incentives to provide usable artifacts.…”