The Center for Data Innovation spoke with Miguel Luengo-Oroz, chief data scientist at UN Global Pulse, an initiative of the United Nations Secretary General focusing on data-driven innovation for sustainable development and humanitarian action. Luengo-Oroz discussed the importance of big data for international development and unique applications for data-driven technologies in public health.
Alexander Kostura: Global Pulse contrasts “big data for development” with traditional development data like survey results and official statistics. What kind of data falls into this category and how can it benefit international development and humanitarian relief efforts?
Miguel Luengo-Oroz: I would say we are not talking about contrasting but rather complementing and integrating traditional data with new data sources. Citizens today—in both developing economies and industrialized ones—we are generating a growing ocean of digital data every minute of every day, just by going about our daily lives. As we use mobile devices to communicate, buy and sell goods, transfer money, and share our lives publicly on social networks, we leave digital trails. Those trails can be used responsibly to design and implement public policies that can benefit everybody, vulnerable populations in particular.
When talking about big data for development, the human centric classification is key, and I am not referring here to the “four Vs”—volume, variety, velocity, veracity—of big data, but rather to the four areas of: what is actively asked of citizens, such as mobile surveys or crowdsourcing campaigns; what people say publicly in the media, from radio to social networks; what people do when using digital services, such as for example aggregated information on population mobility based on mobile usage; and what data is generated by sensors from human activity, for instance measuring pollution levels in the air.
Kostura: As a network of innovation labs, UN Global Pulse has pursued a number of collaborative research projects and data-supported experiments with partners from both the public and private sectors. How have these partnerships and their research results helped to advance your mission of accelerating the use of big data for development?
Luengo-Oroz: Collaboration is key. Since our inception in 2009, we have been researching, innovating, and advocating around many of the principal challenges related to transforming digital data into better outcomes for the poor, discovering new approaches, building tools, and working to demonstrate ways to overcome barriers to adoption and scale. The Global Pulse network of partners and collaborators includes not only UN agencies and government entities, but also private sector companies, nongovernmental organizations, and academia.
These types of collaborations have already enabled big data innovation projects across the UN system—we published 20 case studies last year—but perhaps more importantly, the opportunity is being recognized by many others, and inspiring to help push a movement where new data sharing, privacy, and implementation frameworks are being created to address some of the world’s most pressing issues. The UN’s 2030 Agenda for Sustainable Development will require much more of this type of collaboration for social good. It is fantastic to see how new academic programs are being created to address these challenges, but we need more. My dream is that one day, the UN and the social entrepreneurship sector will be at the forefront of innovation, so the best minds of the new generations work together to solve global challenges that really matter for the future of our planet.
Kostura: In a recent collaboration with the University of Cambridge and the Universal Postal Union, you combined international postal data with other global network data such as trade and airline flights to approximate national indicators of societal well being. Why is it important to test data from both non-digital sources like postal data and digital sources like social media?
Luengo-Oroz: UN Secretary-General Ban Ki-moon has called for a “people’s agenda, a plan of action for ending poverty in all its dimensions, irreversibly, everywhere, and leaving no one behind.” Data coming from non-digital sources like postal flows can be analyzed to shed light on the well-being of countries and populations, especially in those areas where data from digital networks is scarce.
At Global Pulse, we are looking at innovative and groundbreaking projects and tools that can help support sustainable development and humanitarian action at macro and micro level. Take for example the recent tool launched by our lab in Kampala using public radio broadcasts to help identify the priorities of Ugandans. The tool leverages radio as a data source because in the country it is one of the most widespread communication channels that people use. From radio, to postal data, to social media, to mobile money, to remote sensing, there are so many opportunities. The critical thing is to frame the right problem and have a multidisciplinary team that is able to translate real problems—from public health to food security or climate—into data problems. We need many new professionals that can speak both humanitarian and data science languages.
Kostura: Could you explain the concept of data philanthropy? How big of a role does it currently play in supporting international development and relief?
Luengo-Oroz: The truth is that the data revolution has been underway in the private sector for a long time. Most of the data that is out there is held, used, and monetized by private companies. Global Pulse engages with forward thinking private sector companies that have this valuable data, sharing it in ways that fully respect the privacy of their customers, doesn’t compromise their own competitiveness in the market, and yet could be used to make public policies and understand what’s happening while it’s still happening. That is what data philanthropy means to us. At the end of the day, citizens are creating the data, therefore it should be used to protect human rights in all its forms.
Ultimately, we envision a world in which the private sector routinely contributes to a real-time data commons where information on citizens’ wellbeing could be aggregated and shared, as currently happens with weather data.
Kostura: You also founded MalariaSpot.org, which seeks to crowdsource the diagnosis of diseases like malaria in developing countries. How exactly do you crowdsource a disease diagnosis?
Luengo-Oroz: The process of exploring medical images to diagnose disease is time consuming and requires trained specialists. The MalariaSpot idea is to make a video game that encapsulates micro-tasks equivalent to the diagnosis protocol. When people play the game, they are looking for malaria parasites in real blood samples that have been digitalized in the field using a 3D-printed mobile microscope. The analysis of different players around the world is combined into a single collective image diagnosis decision—a crowd-computing methodology that leverages the best of AI and humans. We have found that around 20 players are equivalent to an expert microscopist when looking at malaria samples. The project is still in the prototype phase: last year, we did the pilot with the first ever real-time collaborative diagnosis from Mozambique and so far more than 100,000 people around the world have played and learned how a malaria parasite looks under the microscope.