When a Virus Goes Viral

The Ebola epidemic in West Africa has been in the news lately, and it captured the attention of NYU ITP Master’s student Craig Pickard. Craig was intrigued by the way that the portrayal of the disease by the media influenced reactions to it in the United States, and in particular by the way that social media was able to rapidly disseminate information and connect large groups of people. He decided to design an interface to display exactly what happens when something “goes viral” on social media, and chose to use an actual virus as his focal point.

“If we take a minute to stop and think about what we mean when we say that something on the Internet has ‘gone viral,’ it becomes apparent that a comparison is being drawn between the way in which information spreads and the way an infectious disease might: presumably at an exponential rate. I thought it would be an interesting idea to represent visually how the spread of information, through a network like Twitter, resembles the spread of an infectious disease through the human population.”

Craig built a data visualization program that takes data from Twitter and illustrates how individual Tweets relate to and influence one another:

“I wanted to show the way in which information spreads across a network, likening it to the way a virus might spread from cell to cell in the human body. For this reason, I chose to represent the data I extracted from Twitter as cell-like organisms, moving independently of one another while still forming part of a larger system. Each particle stores data specific to that Tweet, such as the user’s ID, the actual Tweet text, the number of times it’s been retweeted, and a list of the other hashtagged words the Tweet contains.”

To make this possible, Craig used Temboo’s Twitter Choreos and Processing SDK. He designed two processes around the Choreos to drive the visualization: the first imports one hundred recent and popular Tweets that include a specified hashtag (in this case, #ebola) to populate the program, and the second runs every thirty seconds after that to import ten new hashtagged Tweets.

Finally, with the mechanism behind the program up and running, Craig gave his visualization an interactive component:

“I decided to try and simulate the meticulous and calculated feel of a laboratory environment, where movements are small and delicate. I wanted the user to have a feeling that they were physically interacting with the data in much the same way a lab technician would handle live virus samples. To mimic a sterile environment, I wanted the user to have no actual physical contact with the application (further reinforcing the theme of disease and how it spreads). As a result, I decided on using the Leap Motion controller, as it provides a high degree of accuracy, has a well documented library for Processing, and allows for delicate and precise gestures like the pinching of thumb and forefinger.”

Using the Leap Motion controller, viewers of the project can grab individual elements in order to manipulate them and inspect the Tweets they represent.

Although Ebola made for an interesting and topical initial examination, the visualization does just as good a job at displaying the propagation of any hashtag on Twitter. It’s a cool project, and a great way to understand a bit more about how viral content spreads, and how appropriate “viral” is as a descriptor for information spread in the social media age.