In Dhaka, Bangladesh, an estimated 40 percent of the inhabitants live in informal settlements, making an accurate population assessment of a city known for its climate migrant inflow nearly impossible.
To address this challenge, researchers affiliated with the Mobile Data, Environmental Extremes, and Population Project (MDEEP) are using aggregated and anonymized location information culled from mobile call data records (CDR’s) to create large-scale population displacement models to understand population movement related to natural disasters. Though the context in urban Bangladesh is much different than the outskirts of Port-au-Prince, since 2010 there have been a series of case studies which demonstrate the value of leveraging CDR’s for the purposes of disaster risk mitigation.
Recognizing the potential to harness a new form of analysis—not only for crisis response, but for improved planning and preparedness—in August 2013, The Rockefeller Foundation and PopTech co-hosted a convening at the Bellagio Center to better understand the relationship between Big Data and community resilience. Building off the learnings from that convening, the Foundation spent the following year understanding how a conceptual framework would translate into practice, as well as the barriers that prevented many existing projects, such as MDEEP’s, from reaching scale.
The Big Data for Development field is a crowded, diverse, and collaborative space, and still maturing with an ever-increasing range of opportunities emerging. By grouping our projects into two key categories, it allows us to most effectively amplify best practices with an aim towards allowing the field to organically evolve:
1. Local Context is Key
Many of the Big Data for Development projects currently underway have been constructed in a participatory manner, with significant design input from local stakeholders. For example, one of MDEEP’s member organizations, Flowminder, works directly with local governments and mobile providers when it designs CDR analysis projects. Not only does this give a voice to individuals represented through the data, but it builds the capacity of local organizations to lead similar projects on their own in the future.
Within the U.N. system, many agencies have begun exploring ways to leverage big data for their own purposes, all while doing so through co-creation process. The UN Office for the Coordination of Humanitarian Affairs (OCHA) has successfully built a platform to enable the aggregation, opening, and sharing of datasets to advance its mission. As the platform becomes integral to their operations, OCHA recently decided to launch a data lab in Nairobi as a physical space where technologists, the humanitarian community, government, and other actors can all come together to collaborate on the platform to prioritize regional needs.
2. Cross-sectoral Stakeholder Collaboration
Data scientists seldom speak the same language as development practitioners. In comparison to the private sector where there’s often an apparent commercial value, the development community is less incentivized to augment their work with potentially risky technology. Recognizing this disconnect, the Data-Pop Alliance was recently co-founded as a “think-and-do” consortium by the Harvard Humanitarian Initiative, MIT Media Lab, and Overseas Development Institute to bring together the two communities of practice in a more coordinated fashion.
Resilience building requires diverse groups to come together on issues that affect them in the immediate term and the unplanned future. So as we usher in the New Year, let’s ensure that we continue to build capacity for both local and global actors to leverage Big Data in resilience building for the communities in which they live and work.