Analytics and big data are becoming business imperatives. One big element that separates them from unstructured data pools or BI drill-down reports is that analytics can be used to “tell a story,” typically with visualizations.
The visualization angle of this storytelling is typically reinforced with the same types of charts you’ve seen before (you know, the ones that barely register a response during business meetings) or is danced around in attempts to dazzle you with data, data, data.
When I think of art that expresses a narrative, my mind goes to the Albright-Knox Art Gallery in Buffalo, N.Y., specifically the massive work by Anselm Kiefer that hangs in one stairwell. Kiefer’s “The Milky Way” (seen at left) is a few dozen feet of bleak canvass I’ve always interpreted as a destroyed farm, mostly in greys, rusty reds and barren whites. Your mind can go and go with that type of piece, often with all of the joy and smiles the term “destroyed farm” brings along. Another favorite story-teller piece is “Refrigerator Pies” by Wayne Thiebaud, hanging closer to my home at the Milwaukee Art Museum. The title kind of gives it away – it’s a display of pies, as you’d see in a diner, all thick with oil paint portraying cream, crust and cherries – and its directness of its subject quickly fills in the rest. “It’s just pies,” my internal critic says. (Another critic reminds me how much I like to eat pie.) But there’s a prettiness and charm in the presentation of the pies that speak to the power of a straightforward image.
Not to delve too deeply into my C+ college art history course background, but the fluctuation and purpose of these pieces and others came to mind a couple of times this week in discussions on the direction of visualization in the realm of data. After this week, part of me is left thinking there’s a movement of artists-cum-data practitioners that should have those who want to tell great stories with their business data ready to have a few more people wearing berets around the office.
We had lively input Tuesday from the vocal crowd in the Twitter “#IMChat” on data visualization led by my colleagues Julie Langenkamp-Muenkel and Whitney Eden. (To summarize one important lesson from the chat: simplicity = good, donut charts = very, very bad) That same day, I saw in-person displays of the literal practices of data visualization. At Predictive Analytics World in Chicago, Robert Lancaster’s presentation on work in raking and assessing hotel information for travel site Orbitz featured dispersals of dots that gave gravity to the amount of data they were shifting into Hadoop and MapReduce. IBM, in another nail-polishing show of their big data potential, brought an air of humility to its analytic accuracy hits-and-misses in devising “Jeopardy!” champ computer Watson.
Click here for the rest of the article
Source: Information Management