Data Science for Social Good, a data science fellowship personally bankrollled Google chief Eric Schmidt, has commenced for its second summer. The fellowship sees 48 fellows (selected from a pool of 300 candidates) come together “to tackle problems that really matter.” Partners this year include World Bank Group, City of Memphis, and Montgomery County (MD) Public Schools. The fellowship has a focus on making their projects open source after the fellowship, and trying use open source tools where possible to make them cheaper and easier to use for nonprofits.
“One of the problems nonprofits face today is, even if they have the resources and the people’s capability to do this kind of work, they have to start from scratch,” said Rayid Ghani, Obama’s chief data scientist. “There aren’t many existing, easily available, free or cheap tools they can start building on top of, designed for their needs.”
One project this year is attempting to develop a programme which will carry out “text analysis of government spending bills to understand pork spending.” The project will be overseen by DSSG mentor Joe Walsh, a political science PhD candidate from University of Alabama. The project will use machine learning to identify earmarks, which will help to understand where money has been allocated. If successful, the programme could replace the tiresome and ineffective process of trawling through congressional documents by hand, looking for these earmarks.
Although the idea of researchers and students coming together to save the world (or at least improve it) might sound slightly idealistic, there is evidence of last year’s research projects being put to work by nonprofits. One of last year’s fellows, Paul Meinshausen, built a map of the City of Chicago Data Portal to make the city’s data easier to access and more visually streamlined. According to Meinshausen, “Within a day, a couple of Code for America fellows had reused the code to build maps of Boston and San Francisco’s data portals. Jason Lally, an urban planner with Place Matters, substantially improved the way the map collects data about portals and built a map for every Socrata data portal.”
It will be interesting to see what this year’s fellows come up with, and how their research will be used in the wider world.
Interested in more content like this? Sign up to our newsletter, and you wont miss a thing!