You are here:  Home  >  Data Blogs | Information From Enterprise Leaders  >  Current Article

Framing the Data-Science Narrative through Spatial Composition

By   /  March 19, 2015  /  No Comments

by James Kobielus

Data scientists are storytellers who use strictly fact-based numbers, pictures, and words to achieve their desired outcomes. But these data-science “story elements,” as it were, don’t do this simply by jumping onto the screen in any old order. Instead, their arrangement in space, and to a lesser extent, in time, can all the difference in presenting a package with a clear meaning, as opposed to an unholy mess.

Rich visual narratives suffer when they grow too “busy,” a term that refers not just to the density and incongruity of their spatial arrangement but also to the amount of work that needs to be performed by the consumer in order to extract some insight. In practice, what that means is that the target consumer of the data-driven narrative, such as a marketing manager trying to determine whether a campaign truly boosted customer loyalty, must spend far longer to confirm whether that has indeed happened than they would have if the visualization had flagged that finding more saliently.

In a sense, effective visualization is a bit like the ancient Chinese philosophy of “feng shui,” which, stripped of its metaphysical overlay, focuses on the harmonious spatial arrangement of objects in a physical environment. The underlying feng-shui aesthetic is all about achieving a dynamic spatial balance of light, color, textures, and shapes among the physical objects within a house, office, or other built environment in order to facilitate the flow of energy—bodies, behaviors, attentions, tasks, etc.—that inhabits that space.

Energy, in a work context, refers to productivity, and that’s what a spatially harmonious and expertly designed data visualization can deliver. Data visualizations are key productivity tools for knowledge workers. Visualization is key to the productivity of data scientists in their exploration of the correlations, segments, anomalies, and other patterns inherent to the empirical data at their disposal. And those visualizations are also a productivity tool for the executives, middle managers, business analysts, and other knowledge workers engaged in evidence-based decision making. The wrong visualization—whether it be generated by a spreadsheet, a business intelligence tool, a predictive analysis model, or some other data platform—can stop productivity dead in its tracks and intensify the cloud of confusion surrounding the practical matter that the visual would supposed to illuminate.

In order for users to derive maximum productivity from a data-scientific deliverable, it must seamlessly blend the narrative flow of good storytelling, the interactive flow of a well-wrought visual model, and the perceptual flow of a spatially harmonious visualization. For more depth on those topics, check out the following my posts: on the art of non-fictional data-scientific storytelling, the fundamental visualization principles for such storytelling, and the core perceptual principles for visualization design of any sort.

Also check out this very visual and detailed recent article on the art of multilayered data-visualization storytelling through an approach that the author, Giorgia Lupi, refers to as “info-spatial compositions. The key, she says, is that “everything depends on the concept of layering, establishing hierarchies and making them clear,” where “selecting, analyzing, comparing, building hierarchies, etc., is in direct conjunction to the visual development of the layers.”

It’s a very fascinating presentation that any visualization designer, data scientist, or business analyst should immerse themselves in. To me, the feng shui of it all (though they don’t make this connection) is readily apparent in this key statement: “The final fine-tuning of the piece is the necessary effort
required to please readers’ eyes: a well-balanced image where negative space and light elements
play their role aesthetically.”

As with a tastefully arranged living or working space, one immerses oneself in and inhabits the best of these spatially balanced data visualizations. Hopefully, you can also gain some useful insight that you can take away from the immersive visualization and apply to your job.

About the author

James Kobielus, Wikibon, Lead Analyst Jim is Wikibon's Lead Analyst for Data Science, Deep Learning, and Application Development. Previously, Jim was IBM's data science evangelist. He managed IBM's thought leadership, social and influencer marketing programs targeted at developers of big data analytics, machine learning, and cognitive computing applications. Prior to his 5-year stint at IBM, Jim was an analyst at Forrester Research, Current Analysis, and the Burton Group. He is also a prolific blogger, a popular speaker, and a familiar face from his many appearances as an expert on theCUBE and at industry events.

You might also like...

AI Infuses the Next Generation of Web Application Development

Read More →