Big Data

6The Art of Data Visualization

As noted above, data visualization plays a critical role in both understanding data, i.e. in the analyst’s interpretation of the results of the computations made on the data, and in conveying that understanding. As a result, the graphical representation is both a source of intelligibility for the analyst and a rhetorical lever to communicate and convince a recipient. Graphical representation is not the hallmark of big data, and it is by no means a recent invention. As we have amply seen through its role in the exploratory data analysis that emerged in the 1960s, particularly with John Tukey, it predates him by a long way. The cultural sciences, and especially geography, have developed an art of visual materialization that predates the development of computer science, and a fortiori its massification; its uses go far beyond the context of scientific practice. Nevertheless, the density of data and the complexity of the instruments increase the need for visual mediation to understand the results of the processing. This need for mediation takes on specific forms in the context of big data, while at the same time remobilizing or restoring meaning to graphic forms that have more or less fallen into disuse.

6.1. Graphic semiology

In its principle, the service rendered can be described simply: visualization simplifies and summarizes what cannot be embraced with a single glance, be it the “physical” eye or the “mind’s” eye and its intellectual apprehension ...