Innovative maps generated from Big Data could help end poverty.
Researchers use big data to improve poverty maps, a tool which can help identify the most vulnerable quickly and help end poverty.
How do we determine who is extremely poor? Historically, policymakers have depended on surveys and census data to examine and determine extreme poverty.
Though this is an effective technique, it is a resource-intensive procedure which often lacks the detail that governments, relevant organisations and public authorities require in order to intervene suitably.
Researchers are developing new computational tools which use mobile phone records and data from satellites and geographic information systems to generate detailed real-time poverty maps.
“Despite much progress in recent decades, there are still more than 1 billion people worldwide lacking food, shelter and other basic human necessities,” says Neeti Pokhriyal, one of the study’s co-lead authors, and a PhD candidate in the Department of Computer Science and Engineering at the University at Buffalo.
But how do we define what extreme poverty is? Certain organizations define extreme poverty as a critical lack of food, health care, education and other basic needs. Other organizations associate poverty with income; for instance, the World Bank states the extremely poor are those living on less than 1,25 USD a day (2005 prices).
Though poverty is declining in most regions of the world, approximately 1,2 billion people still live in conditions of extreme poverty, mostly in Asia, sub-Saharan Africa and the Caribbean.
The study, called “Combining Disparate Data Sources for Improved Poverty Prediction and Mapping” focuses on Senegal, a sub-Saharan country with a high poverty ratio.
The first data set comprises of anonymous information from eleven billion calls and texts from over nine million Senegalese mobile phone users, which includes how, when, where and with whom people communicate.
The second data set examines the presence of electricity, pavements, agriculture and other signs of development derived from satellite images, geographic information systems and weather stations, thus giving insight into food security, economic activity, accessibility to services and other indicators.
While the current Poverty Maps of Senegal divide the nation in just four separate regions, the researchers combined the two data sets using machine learning, to generate incredibly precise maps specifying the level of poverty of 552 communities within Senegal. The maps are also able to help predict significant aspects such as education deprivation, living standards and health, and can help policymakers to implement enhanced measures to fight poverty.
The maps can be generated very quickly and cost-effectively, and can be updated as often as the data sources are updated, as opposed to surveys and censuses which take years and can be extremely costly to produce. It is anticipated, the method could be used in regions of war and conflict, as well as remote territories.
--
The work is supported by the Bill and Melinda Gates Foundation. The study’s other co-lead author is Damien Jacques, a PhD candidate in the Earth and Life Institute –Environment, Universite Catholique de Louvain, Belgium.
Photo Credit: Courtesy of The University at Buffalo; Creative Commons Attribution Non-Commercial No Derivatives license. Story Source: Materials provided by University at Buffalo. Original written by Cory Nealon. (Content may be edited for style and length)
