Mar21-humanitarian-data
Volunteers will explore existing data on humanitarian response to identify areas of improvement opportunity - e.g. suggesting areas of greatest potential impact to provide care, and proposing optimization of routes for service delivery.
full brief here:
Introduction The consortium of Save the Children, CARE, and Oxfam International represent a significant amount of the humanitarian response across the globe. These three organizations are leaders in the application of technology to development and have each embraced digital and data transformation over the past decade. However, they recognize that sector change - the prioritization and adoption of data-forward activities (digital data collection, analytics, decision-making) - will only come if leading organizations like theirs embrace - and demonstrate - frameworks and systems that showcase the value of data. In this pilot program, the consortium will focus on improving their understanding and implementation of multi-sectoral response to COVID-19 using both organization-held data as well as public data in key service areas.
Problem Save the Children, CARE and Oxfam are leading global NGOs that, together, provide humanitarian support in over 120 countries. While planning and providing that support, these three organizations produce a tremendous amount of data, providing detailed information on everything from financial transactions to fleet management to beneficiary demographic data. Collecting, analyzing and visualizing that data helps these organizations to plan humanitarian response activities and make decisions regarding where best to invest resources so that they can maximize their impact.
Although these three organizations have an increasingly large amount of data available, the underlying environment does not promote a systematic and effective use of data to deliver lifesaving work to disaster-affected communities around the world for a number of reasons, including: Being limited to data that each organization has collected on its own, rather than all the data that has been collected on the country, whether openly available or locked in private databases Having limited access to data science experts who can help to interpret and visualize data in ways that support concrete decision-making
As a result, these three organizations - and the entire humanitarian relief sector - often duplicate resources, serving the same people or communities rather than focusing on complementary services or geographical areas so as to optimize individual resources by collaborating to make the greatest impact possible. To address that issue, Save the Children, CARE and Oxfam have committed to work together in order to share data, identify additional relevant data that lies outside of each individual organization & increase capacity to use data by collaborating with partners such as Datakind. Specifically, this consortium wants to understand if and how public data can be synthesized at a subnational level such that organizations can best direct resources for health and economic intervention.
As a concrete example, to respond to the global COVID-19 pandemic, all three consortium organizations have developed direct health interventions to keep people safe from the disease, and they also continue to provide life-saving services for the world’s most vulnerable populations through WASH interventions, nutrition programs, providing shelter and direct cash/economic voucher programs. By assigning resources based only upon the data each organization has collected separately, however, the consortium is unable to coordinate in order to target both direct & indirect COVID-19 interventions for greatest effectiveness.
Through this DataDive, volunteers will work on this problem in three main ways: Enhance the interoperability and utility of publicly available data to create tools and insights that can be utilized by consortium members along with their internal data to improve targeting of communities in need of support Use multivariate modeling and geospatial analytical techniques to develop multiple targeting approaches with the aim of learning from - and providing - recommendations for increased targeting so that the risks of the pandemic are mitigated for more people. Create new datasets for large scale image detection modeling to create proxy tools for wealth estimation