Bringing social science back to data science.
We are a research team housed within the Annenberg Innovation Lab at the University of Southern California studying social media data and platforms through a mix of social scientific and computational methods. We also build tools and apps for our research and the wider community.
TWITTER SENTIMENT REAL-TIME DASHBOARD
We created a real-time dashboard for collecting and analyzing tweets related to the 2012 Presidential elections.
Dashboard 1, Dashboard 2
TWITTER SENTIMENT HAND-CODING TOOL
Sarcasm is hard to detect with natural language processing algorithms, so we built a tool to code (by hand and/or Mechanical Turk) a real-time stream of tweets. These annotations were then used as features in our algorithms.
Code on Github
TWITTER, HUMOR, and the DEBATES
We analyze how humor plays a role in participation on Twitter during the 2012 Presidential debates, especially regarding content created by highly-visible versus ordinary users.
POP CULTURE, POLITICS, & BLACK TWITTER
Watching engagement on Twitter with television shows like Scandal, we investigate the participatory politics of "Black Twitter" and racial communities on the microblogging platform. Read a white paper summary of our project: [Project Page]
NEWS PRODUCTION IN SOCIAL NEWS SITES
During Hurricane Sandy, the popular social news site Reddit crowdsourced coverage of the crisis. We look at the types of user-generated content that became popular through different processes of content moderation.
DSAIL is led by Francois Bar, Associate Professor of Communication in the Annenberg School for Communication & Journalism. We encourage any interested students to attend our meetings. Email kedrisco-at-usc.edu with questions.
For their generous support of our team, we would like to thank: