Visualizing the structure of the dataviz commumity on Twitter

(see explanations below the visualization)

Visualizing the structure of the dataviz commumity on Twitter

  • The software used for this viz is Gephi, the browser-based version is made with SigmaJs
  • Data were kindly provided by Jeff Clark (@JeffClark) and can be found here.
  • Description of the data: a list of triples: {datavizer A, datavizer B mentioned in the tweets of A, frequency of these mentions}
  • Treatment: I applied a similarity measure "connect two datavizers only if they frequently refer to the same other datavizers C, D, E, ... in their tweets".
  • The result is imported in Gephi where a community detection algorithm was applied.
  • Note that the size of the nodes refers to a centrality measure unrelated to the similarity exercise.
    It is simply: "how many times a datavizer is mentioned by other accounts"

  • This study is part of a larger project: creating a simple Java program to highlight the subtle structures hidden in "hairball" networks.
  • Check out "Gaze", the Github repo for this project.

  • An executable version (exe file) of this program is available here.
  • The Gephi file (with node labels and different attributes) is also available for download

  • Did you like this viz? Follow me on Twitter (here) or check my other projects here