Every day, we don’t give much thought to the tracks we leave behind us on the Web. But other people find these tracks intriguing – as we know all too well since the NSA affair hit the headlines globally. However, tweets and posts also inspire artists, as the show “Big Data Art” in Munich recently proved. Some of data visualizations on show look like a network of blood vessels, others trace the silhouette of Manhattan like lights glowing in the dark. The exhibition included contrasting visualizations of works of art that – in the broadest sense – examine the digital world and how we deal with it.
What is technology doing to us? Do we have the right to be forgotten in the digital world? The aim of the exhibition was to confront the observer with questions like these. For example, at first glance, “Big Data Congress” by the Munich artist Tom Schulhauser, born in 1982, has nothing to do with gathering data on the Internet: it is rather an ensemble of nine different-sized oil paintings of people, including Elvis Presley in a white, flared trouser suit, a man in a heat-resistant suit, and someone lying curled up on the floor. The connection becomes apparent when you take a second look and if you know Schulhauser’s methods. He finds his subjects using Internet searches. The arrangement of the different-sized canvasses in Big Data Congress looks like the results of an image search using Google. In his works, Schulhauser wants to demonstrate the randomness of Internet use. Instead of dwelling on one picture, the observer’s eyes move from one painting to the next.
Big Data in Art: visualizing data flows
In comparison, the representation of data flows cannot actually claim to be art, as Marie-Thérèse Kramer and Sandra Marsch – the two women behind the Munich gallery Munikat – stress. Instead, they are pure visualizations, created by the consultancy d.core with software that is used in a similar way to visualize gene sequences. Using inkjet printers, these Giclée prints are put onto canvas or sometimes in LED light boxes and illustrate, for example, the mobile use of the messaging service WhatsApp over a certain time period, or even pictures containing geo-information uploaded to Flickr in four cities. Our image gallery above shows the data visualizations that were on show. On the following pages, you can read more about the images and what is behind them.
Explanations of the individual images:
- European Travel Patterns: geo-coded tweets
- Mobile Carpet: WhatsApp and Facebook
- Big Data Congress: Google image search revisited
- Medusa: network of Facebook friends
- See something or say something: Flickr and Twitter
- Flickr-Seasons: the changing color spectrum throughout the year
- Edward Snowden: saint or traitor?
- Abstract 447: data tracks rearranged
- Just Coffee
1. European Travel Patterns (photo: d.core): Between three and 10 percent of tweets contain information about where the message was sent from. For this representation, data visualizer Eric Fischer consolidated 750,000 randomly-selected tweets with geo-information from travelers to form an image. It shows the main routes and where the travelers tweet. But even Fischer himself is doubtful about whether the routes pictured are realistic, as he says in a comment on his Flickr profile. He is surprised by the above average use of Twitter in the United Kingdom and the Netherlands. According to Fischer, he retrieved the data using the Twitter streaming API.
2. Mobile Carpet (photo: d.core): The graphic shows almost a month’s use of WhatsApp (red) and Facebook (blue). The data was gathered by the company Arbitron Mobile (now Nielsen Radio), which specializes in such analyses. Each dot represents the use of one of the two apps by a participant in the survey during a 60-minute period. Time is shown on the x-axis from left to right. From top to bottom, the market researchers automatically sorted the participants according to their usage intensity from “heavy users” to “light users.” For the visualization, d.core deployed Big Data software that is based on gene sequencing algorithms.
The random nature of a Google search
3. Big Data Congress by Tom Schulhauser (photo: Tom Schulhauser): The Munich artist Tom Schulhauser finds the main protagonists in his works by searching for generic terms on the Internet. From the results, he filters out the subjects of his paintings. The way in which he arranges the individual paintings is reminiscent of an image search using Google, tempting the observer to “swipe” from one painting to the next, like on a tablet or smartphone. In such a way, Tom Schulhauser aims to expose the indiscriminate nature of surfing the Internet. Instead of taking time to take in an individual painting, the observer’s eyes dart from subject to subject.
Next page: A jellyfish of friends on Facebook
4. Medusa (photo: d.core): This picture shows the network of Facebook friends connected with Benedikt Köhler, a data expert at d.core. An algorithm automatically places contacts with similarities close to each other. This means that different circles such as school friends, colleagues, and relatives can be determined from the picture. The algorithm does not use additional information about the contacts. It sorts them only on the basis of the network structure. The exhibition commentary alludes to the fact that intelligence agencies use similar methods to identify terrorism suspects based on their “friends-of-a-friend” networks.
5. See something or say something (photo: grasundsterne/d.core): On the right, New York with the distinctive silhouette of Manhattan in the center; on the left, San Francisco: For these visualizations, Eric Fischer selected the geo-locations of Flickr photos (in orange) and Twitter posts (in blue). The places at which people both tweeted and took photos are shown in white. Without additional information, the observer can gain an insight into communication behavior.
Four seasons of photos using Flickr
6. Flickr-Seasons (photo: grasundsterne/d.core): For this visualization, d.core retrieved 2,500 randomly-selected photos per month over the course of three years, between August 2010 and July 2013. Each row represents a month and reproduces the color spectrum of the images from this period. Statistically, a high correlation is apparent between the amount of yellow and the average temperature.
Next page: Edward Snowden: Hero or traitor?
7. Edward Snowden by Max Fesl (photo: Max Fesl): The 25-year-old Munich artist Max Fesl painted three portraits especially for the Big Data Art exhibition. As well as the NSA whistleblower Edward Snowden (pictured here), he used the same style to depict the WikiLeaks founder Julian Assange and Glenn Greenwald, the journalist who edited the information on the surveillance program Prism and published it together with an interview with Snowden in The Guardian newspaper. Hanging next to each other, the three works remind the observer of a religious triptych, similar to those found in churches. The association is deliberate, because the three are regarded as both saints and traitors in spying and surveillance scandals. With this alienation, Fesl wants to show how divided public perception is. At the same time, it is noticeable how the three faces remain unmistakable as a result of the constant media attention.
8. Abstract 447 by Stefan Saalfeld (photo: Stefan Saalfeld): The Abstracts series by Stefan Saalfeld is based on data visualizations. However, according to the artist, he “recycles and rearranges” the structures.
Births, marriages, and deaths on Twitter
9. Just Coffee (photo: dcore): Closer examination of this graphic reveals the outline of the continents on a world map. Using the Twitter streaming API interface, d.core created a visual representation of 250,000 geo-coordinated tweets. Each tweet contains the phrase “just coffee”. Similarly, d.core created visualizations using the expressions “just born,” “just married,” and “just died.” This demonstrates how social networks are involved in our entire lives, from the cradle to the grave.