Data science that matters requires bringing together right data, right people

Journalism / Article

Jake Porway is founder and executive director of DataKind. Below he writes about DataKind’s work and the question posed by the Knight News Challenge on Data, How might we make data work for individuals and communities? Winners of the challenge will share in $3 million. Apply at Above: Vision Zero vigil, July 2015. Photo by Miriam Young/DataKind. 

Thanks to the prevalence of smartphones and websites, we now have more data about ourselves and our world than ever before. This information has the potential to reveal critical insights about tough social issues affecting our cities, the environment, our health and more. Data science is a powerful way to extract these insights and inform our decision-making; companies use it all the time to increase their profits. At DataKind, we apply these same techniques and advanced analytics in the service of humanity. It’s an idea that’s on our minds as we consider the question posed by the Knight News Challenge on Data: How might we make data work for individuals and communities?

One thing we’ve learned is that applying data science for social good takes more than finding the right data or the right technical approach; it takes bringing together the right people to define the challenge and co-create solutions.

For example, Vision Zero is an initiative born in Sweden in the 1990s that aims to reduce traffic-related deaths and serious injuries to zero. According to the Vision Zero Network, every year more than 30,000 people are killed in traffic collisions around the U.S. While some call these tragedies “accidents,” the Vision Zero movement advocates that they are entirely preventable.

But what interventions and policy changes are most effective to prevent collisions and keep all road users safe?

To answer this question, we’re partnering with Microsoft and its Tech & Civic Engagement Group and city governments across the U.S. as part of our newest initiative, DataKind Labs.

We’re bringing together newly available data sets including open city data, citizen crowd-sourced data as well as data from private companies to identify patterns and predict where traffic and pedestrian fatalities are likely to occur. We’re also bringing together the right stakeholders to understand the intricacies of the challenge, including local government officials, relevant decision-makers, technical and issue area experts, funders and advocates.

In fact, we’ve already kicked off our work in New York with a brainstorming roundtable of civic tech experts from Microsoft, Transportation Alternatives, and local government officials from the mayor’s office, police department, transportation department and health department. For each city we work with, we’ll use this same collaborative approach, bringing together diverse groups of people to ensure we are being sensitive in our analysis and in creating solutions that will inform cities’ efforts so they can best allocate their resources.

Consider supporting our efforts by joining our team or volunteering your skills and follow along for project updates at #VisionZeroData on Twitter. Meanwhile, as you consider your entries for the Knight News Challenge on Data, keep in mind that applying data science in the service of humanity requires the right data, the right approach and the right team. Moving the needle on the challenges that impact us all takes all of us working together. We look forward to seeing what you come up with. 

Calling all data scientists and volunteers: Join our team!

We’re hiring our first Microsoft Fellow at DataKind Labs to join Erin Akred, our data scientist, and fuel these efforts even further. Apply > If you’re near LA, join us for our upcoming October DataDive to be held at the Teradata PARTNERS conference in Anaheim.  Register >

Knight Foundation is collaborating on the Knight News Challenge on Data with Data & Society and Open Society Foundations. Apply and give feedback on other projects at The challenge closes at 5 p.m. ET Wednesday, Sept. 30. Winners will be announced in January 2016.