Thursday, July 12, 2012

Data collaboration for social good

Governments making more data available to the public could create a new, data-driven society in which we all govern and are governed more effectively. Cities could use the data to become more efficient, governments could improve their shared practices, and both new kinds of citizen engagement and private sector opportunities could take root.

Yet, open data does not immediately mean useable data, and few know what to actually do with the growing wealth of open government data. Contributing this problem is the fact that important stakeholders are often not plugged into what data is available or why—that is, the data scientists who know how to process and analyze data aren't necessarily asking the right questions, while the nonprofits and foundations who ask the right questions don't necessarily know how to process and analyze.

At DataKind, we've started working toward solutions. We were thrilled to support the Safety Data Initiative recently launched by the White House Office of Science Technology and Policy and the US Department of Transportation to hold a DataDive in Chicago in May, a weekend event at which non-profits, data scientists, and government employees worked together to solve social problems with non-profit and open government data.

We've teamed non-profits with data scientists before, but this event added a new element. The White House and the US Department of Transportation had just recently launched, a government portal to an extensive amount of safety data, including data sets on everything from pedestrian safety to food safety. Members of the White House team worked with non-profits ahead of time to bring safety data into their problem definitions. By fostering cross-domain collaboration from the beginning, non-profits could ask their questions of the data, government pointed stakeholders to the "right" data, and data scientists found answers to the non-profits' questions. We invited three high-impact non-profits who could benefit from this work:

  • The Chicago Red Cross wanted to use data to understand where fires were likely to occur. We gave them data on local building safety and fire occurrences, and they built a map showing locations of likely fires and analyzed the demographics of those neighborhoods (shown below).
  • Children's Memorial Hospital wanted to build a database of youth violence prevention programs. We pointed them to data sets on violence in the city, from which they were able to scrape together an automated list of youth violence prevention programs.

  • Enlace wanted to understand the correlations between education and crime in their neighborhood of Little Village. From 311 call data about abandoned buildings, Enlace got their first looks into how neighborhood wellness correlated with their efforts.

More important than the visualizations and insights that came out over the weekend were the collaborations. These micro-communities of non-profits, government officials, and statisticians made for dream teams that were able to utilize government data to solve real problems. Many of the teams continue to work together and will carry the projects beyond the weekend. We are only at the beginning of a revolution in data’s valuable applications for social change. We are heartened to see the government working not to be the authority, but an equal player and a leader in coaxing the value out of open data.

Jake Porway is the Executive Director of DataKind.

No comments: