How the UN want to use Big Data to spot crises

UN Global PulseIn the presentation I embedded below, Robert Kirkpatrick from the UN’s Global Pulse team is talking about how United Nations agencies would like to use big data to search for crises in real-time. It’s a fascinating talk and Kirkpatrick shows how his team has been using data mining techniques to monitor bread prices in Latin America and rice prices in Indonesia.

Data highlights changes in behaviour

Equally interesting is what he would like to do in the future: for example get information showing the streams of money being sent via mobile banking in developing countries.  If the UN could see changes in behaviour, he argues, then  they might be able to spot issues that are in the process of developing.

If, for example, people in a certain region stop saving money to their mobile bank accounts and if at the same time there is a significant increase in money being sent from the capital to that region (presumably by relatives), then this could indicate that there is cash-flow in that region problem which might warrant additional investigation.

It’s an interesting idea, though clearly fraught with all kinds of legal and technical difficulties.

  • The idea of using big data intelligence to spot potential social storms is brilliant. Someone needs to get the highest possible level of attention from the US federal government to build a program around this – with a news – alert dahsboard page for the use of the public. Citizens need to know and see what government and UN diplomats know. In your next writeup on this propose an action plan and who would most likely take steps to move this forward. Thanks!

    • Thanks for your comment, Mike. It certainly is a fascinating and interesting approach. I'll keep a close eye on it and will share additional information if I hear anything.

  • kathryn

    We have many indicators in the field to notify us with market conditions, crops ect….However, the real problem is convincing donors and agencies to act on the advance information and make plans to implement preventative measures…. i.e. famine in Somalia. If that can be solved, then this would be very beneficial.

  • The moral challenge lies in not using the information to manipulate the populations being monitored. There is a tremendous burden posed by the technology/power differential existing between developed and underdeveloped communities. Inherent in this proposal is the potential for the data to be gathered without the permission of the monitored individuals and then to use the data to advance activities that are undertaken without any legitimacy, i.e., an expression of consent on behalf of the monitored. In effect this idea's downsides at best suggests the imposition of a reality unauthorized by the those who would be subject to it. If the idea stipulated a requirement for community consent and a procedure of situational alerting and community leadership involvement prior to any data collection and action then the downside problems may be reduced somewhat.