Interested in politics, social media, and big data? The Brown Institute for Media Innovation, the Data Science Institute, and the Departments of Political Science and Statistics are sponsoring a Hackathon on October 29 and 30 where students and faculty at Columbia will analyze data from the web site voxgov.com to help us better understand the 2016 elections.
voxgov does multiple daily scrapes of government internet sources and makes the information it obtains available in a massive searchable database. The data that voxgov will provide include over 1 million documents from all contested U.S. House, Senate, and gubernatorial races, as well as the presidential race. The documents include social media content from Facebook, Twitter, and YouTube and official campaign press releases.
This data offers the opportunity to study the dynamics of elections in new ways. Social media represents a vast new frontier in how candidates for political office interact with potential voters. Yet we know almost nothing systematically about how candidates and campaigns are using this new mode of interaction, nor do we know much about how social media might be changing the nature of elections in democracies.
One of the goals of the Hackathon is to advance our understanding of the role that social media plays in elections. With data provided by voxgov, teams of students from programs in data science, journalism, political science, and statistics will explore a range of questions in an attempt to better understand the dynamics of the 2016 elections. Potential questions for exploration include fundamental ones regarding amount and variation in social media activity, as well as more complex questions such as how the tone of the race has shifted in response to domestic and international events.
How do the candidates compare in terms of the language they are using? Which candidates have been the most positive and which have been the most negative? Which candidates have been the most issue-oriented and how are they speaking about the issues? Which candidates are driving the conversation and why? What can we learn about the popularity and viability of the candidates through social media activity involving “retweets” and “likes”?
These questions will be addressed using the modern tools of data science, with a key focus on producing stories and visualizations that help to distill an enormous amount of information into forms that are accessible to broad audiences. The collaboration with voxgov promises to be an innovative way to showcase the interplay between journalism, data science, political science, and statistics and to demonstrate how such interdisciplinary efforts can give us new perspectives on political and social phenomena.
If you would like to participate in the Hackathon, please fill out this form.