NEW YORK, 22 June (Office of Information and Communications Technology) — The United Nations announced today that Data for Democracy (D4D) has won the Unite Ideas Internal Displacement Event Tagging and Extraction Clustering Tool (#IDETECT) data challenge.
Data for Democracy is an inclusive community-driven initiative for data scientists and technologists to volunteer and collaborate on projects that make a positive impact on society. The D4D #IDETECT solution was built by a team of volunteers from around the world, including Aneel Nazareth, George Richardson, Simon Bedford, Wendy Mak, James Allen, Yane Frenski, Domingo Hui, Charles Neiswender, Daniel Forsyth, Joshua Arnold and Alex Rich.
The #IDETECT challenge, a collaboration between the United Nations Office of Information and Communications Technology and the Internal Displacement Monitoring Centre in Geneva, challenged the public to create an open-source tool capable of estimating the number of people reported as internally displaced around the world, and their locations, with a significant level of accuracy.
Every day, thousands of people are forced to flee their homes — yet remain within their home countries — as a result of conflict, violence and disaster. Many of these internally displaced people remain invisible, making it more difficult to help them meet their basic needs. The Office of Information and Communications Technology and the Internal Displacement Monitoring Centre launched the #IDETECT challenge to call upon the global community of developers for analysis of big data to detect disaster- and conflict-related displacement reported in the news and on social media.
“This challenge demonstrated the potential for open-source technology to significantly advance our understanding of a complex issue with an ever-changing landscape,” said Salem Avan, Director of the Global Services Division in the Office of Information and Communications Technology. “Open-source developers are proving to be an invaluable resource to the United Nations and our partners for efficient, affordable and high-quality solutions that can accelerate our humanitarian response.”
The D4D team’s winning solution uses an artificial intelligence algorithm to predict the number and locations of internally displaced people through URL tagging and word frequency by drawing on a vast amount of online data. Judges deemed the solution to have high potential because artificial intelligence algorithms have the ability to learn every time they are exposed to new data, and that will always be the case when predicting internally displaced people, situations that are often both complex and highly fluid.
“We are very proud to have been a part of the #IDETECT challenge,” said the D4D team. “Monitoring displacement is an inherently data-oriented problem, but not one that can be solved with out-of-the-box solutions. There is clearly great potential for natural language processing methods and machine learning, such as those that we have deployed, to help understand and make links between evolving situations on the ground.”
#IDETECT judges were impressed by the overall quality of submissions to the challenge. In addition to the D4D solution, Honourable Mention was awarded to Samuel Bollier, a master’s student in international relations at Tufts University, for his submission “Human Displacement Analyser”. The Internal Displacement Monitoring Centre will implement both Mr. Bollier’s and D4D’s solutions, both of which will be featured on the websites of Unite Ideas and the Internal Displacement Monitoring Centre.
“Working with the Unite Ideas platform and team enabled the Internal Displacement Monitoring Centre to leverage the efforts of brilliant, talented data scientists from around the world,” said Justin Ginnetti, Head of the Data and Analysis Department at the Internal Displacement Monitoring Centre. “Thanks to Unite Ideas, we were able to move much more quickly than if we had developed the tool ourselves. Using this platform also gave us the opportunity to call attention to the issue of internal displacement and introduce it, and our work, to a community that we had not previously engaged with. #IDETECT is an important step towards enabling [Internal Displacement Monitoring Centre] to exploit the full potential of new technologies,” he added. “In today’s world, many internally displaced people remain invisible; technology can help ensure that no one is left behind.”
#IDETECT is the seventh challenge issued by Unite Ideas, a big data crowd-sourcing platform developed by the Office of Information and Communications Technology to facilitate collaboration among academia, civil society and United Nations offices, and to mobilize data scientists and software developers around the world to help tackle the complex issues faced by the United Nations and its Member States through the creation of open-source technology solutions.
The winning #IDETECT solutions can be viewed at unite.un.org/ideas.