The US political and online news systems have been increasingly defined by their polarization. Since 2016, media researchers and journalists have found strong evidence of this. This has led to two sets of media that speak about completely differing things.
This site lets you explore media polarization via the top words used in article headlines across two sets of American online news sources. These partisan sets are based on sharing patterns of Twitter users, and thus more of an assessment of what registered voters from each party are sharing rather than a direct assessment of the editorial positions of the publication itself. The data is collected from an under-construction new archive of online news being collaboratively built by the Media Cloud organization and the Internet Archive's Wayback Machine.