Dear friends,
The world is not what it was just a month ago. Putin’s troops have invaded Ukraine, civilians are suffering and dying under bombardment, many are fleeing their homes and the horrifying prospect of a large-scale war in Europe frighten people all over the world.
Nobody was expecting it. Nobody could believe that such an event can happen in a stable and developed Europe, but the reality has now taught us a lesson to the contrary… In times like these, the world needs even more solidarity, camaraderie, and altruism. And in fact, Europeans are solidly united as it rarely was the case in the history: people from Eastern and Western Europe are collecting humanitarian aid in cash and kind, volunteers are driving to and helping at the Ukrainian borders, and many people are hosting refugees or trying to help them in a variety of other ways.
This kind of direct help and actions is extremely urgent and necessary! But our actions in humanitarian crises like this could also benefit from the help of intelligent algorithms. Such algorithms are usually not applied at scale yet, but there is a lot of research, prototypes and MVPs of intelligent applications that can help ease the miseries of many refugees. So, we decided to dedicate this newsletter to such applications of Machine Intelligence (MI).
These applications were mostly developed as reaction to other crises, and it will take some time before they will be adapted for the current conflict. We hope that this newsletter will contribute to fostering the usage of MI applications for refugees and will increase awareness of authorities and humanitarian aid organizations about their potential.
But before we start, we would like to draw your attention to the following two organizations that are providing extremely vital help during this senseless war:
Aktion Deutschland Hilft is a union of many German and international NPOs that have joined their forces to collectively provide emergency aid for Ukraine (and other regions suffering under crises).
Unterkunft Ukraine is an alliance of mission-driven for- and not-for-profit organizations that help to connect refugees with people who have free beds for them. If you are blessed with a big enough house or appartement, you might consider registering and offering overnight accommodation for people from Ukraine at your place.
Let us together make the world a better place for all of us.
Your MI4People Team
MI application for refugees
Predicting Migration Flows
So far, in the current crisis in Ukraine, EU countries are acting fast and in a very coordinated fashion. However, in 2015, when refugees from Syria came to Europe many countries were not prepared for the large number of refugees and their humanitarian infra-structures were clearly overwhelmed. Machine Intelligence technologies might help authorities in such situations by predicting the probability and volume of possible displacement in a particular region so that authorities (and also humanitarian aid organizations) can prepare to a new wave of refugees.
Such predictions can be made by analyzing various data sources with Machine Learning (ML) and Artificial Intelligence. Data sources might be very different: from classical censuses, surveys, and administrative data to satellite pictures and latest news articles. For analyzing each of these data sources various ML techniques must be applied, however, they can also be combined to increase the accuracy of predictions.
There are several research groups and projects working on such early warning systems. To name a few of them:
Internal Displacement Event Tagging Extraction and Clustering Tool (IDETECT) is an AI-based real-time platform that automatically monitors news to identify incidents with a likelihood to increase the displacement, e.g., violent conflicts and natural hazards. IDETECT reads the world’s news, reports from UN and various NGOs, social media and other sources of information, filters it for articles relevant to displacement, and extracts key information, such as causes, locations and number of displacements.
Project Jetson is an experimental project launched by United Nations High Commissioner for Refugees (UNHCR) to better understand the reasons that cause people to flee their homes in Somalia. Based on ten key variables of forced displacement, such as commodity market prices, rainfall, and violent conflicts, Jetson uses ML to calculate an index that allows for short-term predictions of expected migration flows out of Somalia.
PulseSatellite is a collaborative web-based tool that uses AI to extract the most relevant information from satellite imagery for use in humanitarian contexts. This system can be applied for various use cases such as identifying the direct impact of natural disasters, and damage assessment, but also for monitoring population displacement and settlement mapping of refugee camps.
Free Legal Aid
On their escape from crisis regions, refugees struggle with many problems. One of these problems is that they usually do not know the laws and regulations for refugees in the destination countries and the process of application for asylum might become an insurmountable hurdle, often language barrier adding to the challenge.
This is where “the world’s first robot lawyer” from DoNotPay provides vital help. The Facebook-Messenger-based chatbot that was originally created to help overturn parking fines and support vulnerable people to apply for emergency housing is now also able to help refugees fill in an immigration application in the US and Canada, and – for refugees in the UK – it also helps with applying for asylum support.
This is not a trivial task, and it took developers and domain experts a lot of work to ensure that the chatbot provides correct legal advice. We hope that similar initiatives and services will be started soon also in other countries all over the world.
Psychological Support
After days and weeks of struggles, anxiety, and deprivation many refugees need psychological and emotional help that is typically not available in overcrowded refugee camps. To ease this problem, intelligent chatbots that simulate a conversation with a friend and provide first emotional help can be used.
One example of such a chatbot is Karim that was developed by AI-company X2AI to support Syrian refugees in Lebanon in 2016. With the help of a non-governmental organization Field Innovation Team, X2AI was able to provide free and on-demand emotional support, cognitive-behavioral-therapy-based interventions, solution-focused strategies, and relaxation techniques in Arabic.
MI-technology behind this application is called Natural Language Processing/Understanding (NLP/NLU) and has experienced incredible advance in the last few years. Extremely sophisticated and complex language models like BERT (Google) or GPT-3 (OpenAI/Microsoft) are (partially) open-source and free to use and can be applied to create application such as Karim with relatively small effort. We hope to see new innovative solutions in this domain for many languages and regions.
Helping refugees to integrate into the new society
When refugees have arrived at their destination country, the hosting countries should try to integrate them into the new society quickly. On one hand, refugees benefit from it both psychologically and socioeconomically. On the other hand, the hosting country can ensure that the refugees fruitfully contribute to the society and can prevent negative developments such as increase in criminality, marginalization, and racism.
However, often it is difficult to find a place where a particular refugee has the highest chance for integration since it is a non-trivial task to match the local capacities and priorities, local employment market situation, refugee’s background, skills, and preferences etc. MI can also contribute to mastering this problem!
An AI-based software Annie™ MOORE (Matching and Outcome Optimization for Refugee Empowerment) was developed by a collaboration of Swedish, UK, and US researchers for replacement agencies in order to help find the best matches between refugees and local communities. Since 2018 Annie™ MOORE has been helping a US resettlement agency and you can read more about the results in the corresponding paper and in a The Atlantic article. Unfortunately, we could not find any recent activities of this project but hope that Annie™ as well as other similar initiatives will be able to support refugees in the current and future crises.
Concluding Remarks
In the current Russian invasion of Ukraine situation, it is even more clear how much suffering and misery exist in this world. We should mobilize all our resources and use every opportunity to fight them. One of the biggest strengths of the developed world is technology, especially in the IT and MI sector. So, it is very logical to apply this strength to help people in need. Thus, think creatively and let us use the power of MI together to make the world a better place for all of us!
Comments