Class Projects - Spring 2013


onMessage helps voters and journalists see the issues politicians are most focused on and how their focus changes over time.

Screencast of the onMessage demo from the Collaborative Innovation class Spring 2013.
What it does

OnMessage analyzes political speeches in an effort to show voters the issues their political leaders are most concerned with. The system looks for keywords in political speeches to determine how often politicians talk about different policy areas, such as foreign affairs, education, economics and health care. It then shows users a chart to show focus and changes in focus over time.

How it works

The crux of this technology is accurately defining “key terms.” The build team selected key terms from presidential speeches, with specific attention paid to "focus sentences" — those that made a clear reference a particular hot-button issue. These terms were collected and added to a dictionary of "mentions." Those mentions are then correlated to a broader category (the environment, for example) and plotted on a timeline based on user-specific searches.

Next Steps
  • Account for terminology not just from speeches, but every type of media campaign message outlet, including social media, interviews and campaign websites. 
  • Create a variation of the tool where a candidate’s actions — their policy work in a given area, for example — are analyzed as well to provide a larger, composite picture. 

onMessage creates visualizations to help users see trends over time.

Overview of the technology.


More about how onMessage was built

Code repository

Initial Concept: Rich Gordon

Student Team: Tara Kadioglu (journalism), Mary Posani (journalism), Cassie Rommel (computer science)

Faculty Guidance: Larry Birnbaum, Rich Gordon, and Kris Hammond (with assistance from Shawn O’Banion)