EXPERIMENTS

(And in-progress projects)

Experimentation is a distinct type of activity within the innovation process. Core to experimenting is trying things out, and using data to inform progress and iterations, prior to rolling out finalised policies, products, or processes more widely. Systemic testing allows us to better understand what may work or what won’t, and allows us to make smaller mistakes that we can learn from early on, rather than needing to make more substantial – and costly – corrections later on. We support experiments and new ways of working with colleagues in Headquarters and field level. We create safe spaces to experiment on everything from new ways of funding, to procurement processes, through to the adoption of new technologies. Experimentation at UNHCR is about embracing risk, testing things out, and learning what works for whom, and where.

JETSON

Predictive Analytics Engine

Jetson is a predictive analytics platform aimed at providing better data to deliver better decisions. We measure multiple variables to see how changes may effect internally displaced populations. Jetson is an experimental project launched by UNHCR’s Innovation Service in 2017 to better understand how data can be used to predict movements of people in Sub-Saharan Africa, particularly in the Horn of Africa.

The Predictive Analytics Engine (Jetson) is an applied predictive analytics project taking concrete steps to provide insights on the future of displacement.

The project is the first step in understanding data about a) the population flow and b) some of the most common variables that are correlated with the population flow. The project focused on the catalysts that might cause people to flee their homes in the Somalia situation. The main objective is to make predictions about potential displacement events by utilizing data mining, statistics, modeling, machine learning, and artificial intelligence to analyse different data and yield some preliminary conclusions.

Video: Insights from the project’s UX Designer on the challenges and the importance of predictive analytics in the humanitarian sector.

SOCIAL MEDIA MONITORING

Quantifying sentiment – Xenophobia in Europe

The Innovation Service with the support of UN Global Pulse commenced a big data sentiment analysis with the Regional Bureau for Europe during the Europe Refugee Crisis with the aim of providing decision makers with additional context to the situation as it unfolded within Europe. Part of the exercise was simply to work out how big data – in this case, social media data – could be used to ameliorate UNHCR’s understanding of a complex, and unique situation. The experiment continues to unfold, taking a range of shorter and longer term events into account.

The monitor was set up within Foresight tool of Crimson Hexagon platform for understanding collective opinion, feelings, statements and their changes over time, identifying patterns of sentiment derived from them particularly of publicly available information the Twitter social media platform. The purpose of this research is to better understand host community sentiment in light of global events so UNHCR can have more timely decision-making driven by big data analysis.

LIVE UPDATES

Real time updates from the field

Reporting in emergencies can be particularly difficult for colleagues working in hard to reach and remote locations. While media outlets may report on the high-level information, we wanted to explore what more informal and live reporting could look like from the UNHCR perspective. One of UNHCR’s Innovation Fellows will be testing this method over the coming months to help address the lack of understanding for staff and refugees in an emergency context and bring insights into the day-to-day work of a humanitarian. Stay tuned for the launch of this exciting project.