Predictive tool to help NGOs better manage migrant needs
For many irregular migrants arriving in the EU, critical points of contact are often organisations dedicated to ensuring their welfare. This work can involve ensuring adequate resources are available, or providing information on the number of asylum applications forecast for a specific EU country. These NGOs however face a range of challenges, not least in anticipating when and where migrants will arrive. “One reason for this is that the information NGOs receive about migrant arrivals is so fragmented,” explains ITFLOWS project coordinator Cristina Blasi Casagran from the Autonomous University of Barcelona in Spain. “This can make it difficult to best allocate resources for future arrivals. Information is not always regularly updated, and it can be difficult to ascertain which sources are most reliable.”
Mapping all sources of information
The aim of the ITFLOWS project was to provide intelligence support for these NGOs. By collecting relevant data concerning migrant flows, and presenting this information in a more accessible and user-friendly manner, the project team hoped to give NGOs a valuable tool to help them better anticipate and prepare for arrivals. The project team began by mapping all available sources of information. This included data provided by the likes of Eurostat and the UNHCR, as well as less well-known sources such as GDELT. “GDELT collects news headlines from all over the world,” adds Blasi. “You can select keywords and filter topics, which could bring new events. For example a revolution in one country could lead to people fleeing.”
Models to help NGOs forecast migrant arrivals
From this initial analysis, the project team developed the EUMigraTool, designed specifically for the needs of these NGOs. This online tool enables users to forecast migrant arrivals, using four predictive models that utilise the data gathered. The first model is designed to help NGOs make rough short-term predictions. Users identify the main countries of origin of migrants, and then access data that enables them to see what is going on. “This can help them to make predictions one month ahead,” says Blasi. A second model zooms in closer to these countries of origin, helping users to identify where, say, new conflicts might have arisen. Blasi points out that most migrants fleeing a conflict zone stay in their home country, or head to a neighbouring country. “Another model provides predictions of migrant flows three to five months ahead,” remarks Blasi. “For this we mostly used Google Trends. By using keywords, we can see what potential migrants are searching for.” A fourth model gives users insights into the perceptions of migrant arrivals among recipient communities. Accessible data includes social media posts that can be classified as negative, positive or neutral.
Recommendations on key migration issues
The EUMigraTool was piloted in Greece, Italy and Spain, and by 20 NGOs, and was well received. The tool is still in the prototype phase, and further funding is required to expand its use. “We were clear from the beginning that this tool is specifically for NGOs and not governments,” notes Blasi. “We do not intend to charge a fee for this service, as many of these organisations rely on donations.” The project team did however prepare a number of policy briefs for decision makers. These provide recommendations on issues such as effective policies for the socio-economic integration of migrants to the EU, and some of the challenges in forming external partnerships to address migration.
Keywords
ITFLOWS, humanitarian, asylum, migrant, Eurostat, UNHCR, migration
