Project: Inferring Social Networks

In this project, I worked in a team with three other people to create visualizations of social networks of public figures. Our software gathered publicly available data, wrote it to a noSQL database, calculated co-location frequencies, and based on those frequencies constructed the visualizations below.

Data sources include: twitter, wikipedia, the NYTimes campaign finance API, syndicated news articles, and high-profile political blogs.

Here is a graph of Hillary Clinton’s network:

And here are two word clouds of Hillary Clinton’s network. The first is based on wikipedia sources:

The second word cloud is based on campaign finance sources:

 

You can see that the different sources produce different word clouds. The Wikipedia cloud links Hillary Clinton to more historical political figures. The campaign finance cloud links her more to successful political operatives.