We collected 90,811 political blog posts over a ten-year period from 2005 to 2015 from seven different sources, including AMERICAblog, Huffington Post, and ProPublica. We then ran text analysis on these blogs and generated 418,641 terms that were classified into four different categories. These terms were then input into TimeArcs. We first computed the sudden attention measure (see Section 4.1 in the paper) for each term and then to computed the relationships between the top 1,000 terms. We filtered relationships of strength at least 17 (i.e., terms that were mentioned together in at least 17 blogs in one month). Finally, the top 200 terms were plotted in the layout depicted in the following figure. This layout provides an overview of major political events in the past 10 years in one display. Please try online demo.

Viewers can lens into a period of time to see the details by mousing over the time axis. The following figure depicts lensing into 2013.