Seminar: Political Behavior in Online Networks

24/12/2019 14:00

In this talk, I will present two recent projects studying political behavior in online networks. The first project studies anger contagion in protest networks on Twitter. Large political protests in the real-world are often reflected in numerous discussions on social media platforms. We ask what makes the anger persist (or not) during protests and what is the role of updating strategies and network positions of different users. We examine the influence dynamics in online social networks and the distribution of anger during protests in Charlotte, North Carolina (2016) and Charlottesville, Virginia (2017). We extract mention and retweet networks of users tweeting during these protests, and conduct emotion analysis on tweets to determine the anger level of the users. Although these two protests differ in their triggering points, they have similarities in their macro behaviors during the peak protest times. Anger tends to peak early on and then taper off. The angriest users have fewer followers than less angry users. We further study the dynamics of anger using computational models implemented on the extracted user networks. This research has the potential to help understand influence dynamics in protest-related discussions and provide insights for public policy implications. The second project studies the dynamics of commenters' networks across time and political spectrum. Millions of people comment daily on current events on a variety of platforms ranging from diverse social media to the news sites themselves. Reading others' comments can shape one's own opinions about the story, its author, and the media outlet and help spread opinions and claims which counter the mainstream narrative. Here we investigate the social network of commenters on four well-known U.S. news sites that span the political spectrum from left to right. We focus on two factors that could affect the dynamics of influence in commenters' social networks: political orientation and threat experienced by one's group. Our results suggest that commenters on political sites on the extremes of the political spectrum tend to have a higher inequality of influence than commenters on more moderate sites. Furthermore, we find the tendency for higher inequality of influence at times when supporters of a particular site feel threatened. This research can reveal whether discussions are influenced by relatively few commenters, or whether they are shaped by many different commenters and thus more in line with Habermasian ideal of public discourse.