Twitter Conversation Trees is a work of data art created by design technologist Peter Beshai to accompany a presentation for Deb Roy and Bridgit Mendler. The work seeks to visualize the toxicity in conversations on twitter. A twitter post is modeled in real time as a cone shaped tree, which branches out from the replies. The more toxic a reply thread is, the more withered it’s branch appears. The data is extracted from popular public posts, using methodology and criteria developed by Bridgit Mendler. In order to generate the visual Beshai primarily utilized Houdini, which meshed well with the generative nature of the project and his background in programing. Furthermore Houdini had excellent python compatibility which complemented the use of the Python library Tulip to structure the trees.

The author takes a fairly neutral stance on the issue on internet toxicity in this piece. The piece conveys the amount of toxicity on twitter, without making a judgement on what one’s reaction should be. This is likely because the presentation this was displayed in was done with scientific intention and therefore had little intent on passing judgement, instead focusing more on the methodology used to register toxicity in a post. While the presentation did, the artwork itself does not display the actual contents of the threads rendered toxic. Instead it chooses to focus on the patterns and prevalence of toxicity in general.

link: https://medium.com/cortico/visualizing-toxicity-in-twitter-conversations-3cd336e5db81