Doing Data Science for Social Good

by Hayley Baldwin - 11 Nov 2015

We recently teamed up with the Data Science for Social Good Berlin (DSSG) chapter to host their first-ever DataDive in our Shuttle Innovation Lab. More than 60 data scientists, data enthusiasts and other technologists from the Berlin tech community spent Friday night through Sunday afternoon working with datasets provided by the nonprofit organizations streetfootballworld, Jambo Bukoba and DataLook. “Data Divers” represented diverse skillsets and backgrounds (eg. NHST, linear models, data visualization, NLP, sentiment analysis, and more), chose their own projects to work on and developed their own workflows. The event also held spots for “observers” — myself included.

Have you heard of DataKind? Well, DSSG Berlin works similarly. They provide data to help nonprofits prove that their programs are effective, and predict trends for future programs. In other words, they put the numbers behind the narratives for social organizations that don’t have data scientists or other technologists on staff. DSSG Berlin linked up to us via Zalando product specialist Kevin Wong, who’s an ambassador to the group and was helping them to find an event location. As it turns out, many of our 60+ data scientists were already volunteering for the initiative in their free time.

In addition to data-hacking, the dive included final presentations and breakout sessions. Zalando Data Scientist Katharina Rasch delivered one such session about her participation in the Data Science for Social Good Fellowship at the University of Chicago this past summer. Katharina was invited to spend three months working with 41 other fellows, project managers, technical mentors and the organizers on data science projects for nonprofits and government organizations.

“I had a grand time this summer” she said, “and I am excited to see that the idea of using our skills for social good is also taking off in Germany. You can learn more about the projects we worked on here.”

Volunteer DSSG Ambassadors helped the data divers manage their time and work flows, and overcome problem-oriented roadblocks. The non-profits’ founders and team members also provided support and helped participants navigate the data (jump to 51:02 in the video below for a very heartfelt thank you from Jambo Bukoba founder Clement Mulokozi)

Watch the final presentations of the results here:

Written Results:

*Note: not all data has been linked yet.

The DataLook team livestreamed the event, and a truly awesome video of their findings that you can watch here.

You can learn more about DSSG’s past and future projects on their GitHub page. A big thanks to the DSSG Berlin team and their volunteers for taking the time to put on such a great event — and for including us! Go here for event photos.

Similar blog posts