Syndicate content

Sustainable Communities

Demystifying machine learning for disaster risk management

Giuseppe Molinario's picture
Also available in: العربية | Español | Français

To some, artificial intelligence is a mysterious term that sparks thoughts of robots and supercomputers. But the truth is machine learning algorithms and their applications, while potentially mathematically complex, are relatively simple to understand. Disaster risk management (DRM) and resilience professionals are, in fact, increasingly using machine learning algorithms to collect better data about risk and vulnerability, make more informed decisions, and, ultimately, save lives.

Artificial intelligence (AI) and machine learning (ML) are used synonymously, but there are broader implications to artificial intelligence than to machine learning. Artificial (General) Intelligence evokes images of Terminator-like dystopian futures, but in reality, what we have now and will have for a long time is simply computers learning from data in autonomous or semi-autonomous ways, in a process known as machine learning.

The Global Facility for Disaster Reduction and Recovery (GFDRR)’s Machine Learning for Disaster Risk Management Guidance Note clarifies and demystifies the confusion around concepts of machine learning and artificial intelligence. Some specific case-studies showing the applications of ML for DRM are illustrated and emphasized. The Guidance Note is useful across the board to a variety of stakeholders, ranging from disaster risk management practitioners in the field to risk data specialists to anyone else curious about this field of computer science.

Machine learning in the field

In one case study, drone and street-level imagery were fed to machine learning algorithms to automatically detect “soft-story” buildings or those most likely to collapse in an earthquake. The project was developed by the World Bank’s Geospatial Operations Support Team (GOST) in Guatemala City, and is just one of many applications where large amounts of data, processed with machine learning, can have very tangible and consequential impacts on saving lives and property in disasters.

The map above illustrates the “Rapid Housing Quality Assessment”, in which the agreement between ML-identified soft-story buildings, and those identified by experts is shown (Sarah Antos/GOST).

World Bank engagement through the Expert Group on Refugee and IDP Statistics (EGRIS)

Emi Suzuki's picture
Better data, based on guidance from the Expert Group on Refugee and IDP Statistics, will help improve our support for the
displaced and their host communities.
Credit: Chisako Fukuda/World Bank

The record-high number of forcibly displaced people today—refugees, asylum seekers and internally displaced persons (IDPs)—has underscored the need to improve the way the global community addresses these situations. The new global compact on refugees adopted at the UN General Assembly on December 17th will guide these efforts.

It is widely acknowledged that statistics are critical to inform our response, but until recently, there were no global standards. Lacking international guidance, different institutions produced data on forced displacement without due coordination or transparency. Terminology was inconsistent, making data incomparable. Statistical capacity varies between countries, and refugees and asylum seekers were not included in national censuses or regular migration and population statistics.

Mapping for sustainable development: The Open Data for Resilience Mapathon

Lorenzo Piccio's picture

On Wednesday, November 14, we joined more than 170 volunteers at the World Bank’s Washington, D.C. headquarters to draw little red boxes on a map of Alajo—a small town in the coastal metropolis of Accra, Ghana.

Some might find tracing a map of a city 8,500 kilometers away to be a surprising way to spend an afternoon, but there are good reasons for it. The boxes represented buildings, and they will go on to become invaluable geospatial data that will help the residents of Accra prepare for and respond to flood risk. Home to over two million people, Ghana’s capital city is highly vulnerable to flooding. In 2015, torrential rainfall left much of the city underwater—affecting 53,000 people and causing an estimated US$100 million in damages.

In just a little over two hours, the volunteers made over 3,000 edits to the map of Alajo, complementing the work of local teams in Ghana that are leading data collection efforts in the field. Once validated by more experienced mappers, the data collected will help guide improvements to Accra’s solid waste disposal management system, and also inform the upgrading of settlements vulnerable to flooding.

Open Cities Africa: Collaborative mapping to build resilient societies

Checklist: 10 guiding principles for effective use of risk data

Simone Balog-Way's picture
Local city officials and university students in Can Tho, Vietnam
collaborate and learn about innovative mapping technology
Photo credit: Robert Banick/GFDRR

Effective decision-making in disaster risk management requires good risk data. That’s why at the Global Facility for Disaster Reduction and Recovery (GFDRR)’s Open Data for Resilience Initiative (OpenDRI), our work focuses on improving processes surrounding the dissemination, creation, and communication of risk data—from using drones to map flood vulnerability in Niger to building a geospatial data sharing platform in Bangladesh.

And while much more progress is needed to improve the quality and availability of risk data, the good news is that governments, international agencies, and scientific institutions are increasingly making their data open and available to planners, civil contingency managers, and responders. Combined with advances in technology, the movement for open data is generating an unprecedented volume of risk data. OpenDRI’s Open Data for Resilience Index monitors this trend by tracking the existence, availability, and openness of data on disaster risk and resilience worldwide.

One key challenge now is how best to capture, analyze, and communicate this data to inform decision-making. In an effort to provide a framework to guide the use of data in disaster risk management, OpenDRI has developed 10 principles that can be applied throughout a project’s life cycle to help ensure that risk data is used effectively for decision-making. Below, we break down these guiding principles and provide practical examples of how they have been applied.

  1. Put users at the center of project design

Risk information must be grounded in the needs of users at relevant geographic and time scales and provided through accessible and understandable formats. In a successful example of this practice, UNDP Myanmar’s SESAME (Specialized Expert System for Agro-Meteorological Early Warning) drew on local cropping practices to develop location-specific agro-advisories which covered multiple timescales.

The 2018 Atlas of Sustainable Development Goals: an all-new visual guide to data and development

World Bank Data Team's picture
Also available in: Español | العربية | Français
Download PDF (30Mb) / View Online

“The World Bank is one of the world’s largest producers of development data and research. But our responsibility does not stop with making these global public goods available; we need to make them understandable to a general audience.

When both the public and policy makers share an evidence-based view of the world, real advances in social and economic development, such as achieving the Sustainable Development Goals (SDGs), become possible.” - Shanta Devarajan

We’re pleased to release the 2018 Atlas of Sustainable Development Goals. With over 180 maps and charts, the new publication shows the progress societies are making towards the 17 SDGs.

It’s filled with annotated data visualizations, which can be reproducibly built from source code and data. You can view the SDG Atlas online, download the PDF publication (30Mb), and access the data and source code behind the figures.

This Atlas would not be possible without the efforts of statisticians and data scientists working in national and international agencies around the world. It is produced in collaboration with the professionals across the World Bank’s data and research groups, and our sectoral global practices.
 

Trends and analysis for the 17 SDGs

Chart: Why Are Women Restricted From Working?

Tariq Khokhar's picture
Also available in: العربية | Français | Español | 中文

Economies grow faster when more women work, but in every region of the world, restrictions exist on women’s employment. The 2018 edition of Women Business and the Law examines 189 economies and finds that in 104 of them, women face some kind of restriction. 30% of economies restrict women from working in jobs deemed hazardous, arduous or morally inappropriate; 40% restrict women from working in certain industries, and 15% restrict women from working at night.

 

Your Cow, Plant, Fridge and Elevator Can Talk to You (But Your Kids Still Won’t!)

Raka Banerjee's picture
Download the Report

The Internet of Things (IoT) heralds a new world in which everything (well, almost everything) can now talk to you, through a combination of sensors and analytics. Cows can tell you when they’d like to be milked or when they’re sick, plants can tell you about their soil conditions and light frequency, your fridge can tell you when your food is going bad (and order you a new carton of milk), and your elevator can tell you how well it’s functioning.

At the World Bank, we’re looking at all these things (Things?) from a development angle. That’s the basis behind the new report, “Internet of Things: The New Government to Business Platform”, which focuses on how the Internet of Things can help governments deliver services better. The report looks at the ways that some cities have begun using IoT, and considers how governments can harness its benefits while minimizing potential risks and problems.

In short, it’s still the Wild West in terms of IoT and governments. The report found lots of IoT-related initiatives (lamppost sensors for measuring pollution, real-time transit updates through GPS devices, sensors for measuring volumes in garbage bins), but almost no scaled applications. Part of the story has to do with data – governments are still struggling how to collect and manage the vast quantities of data associated with IoT, and issues of data access and valuation also pose problems.

Announcing Funding for 12 Development Data Innovation Projects

World Bank Data Team's picture
Also available in: Français | 中文

We’re pleased to announce support for 12 projects which seek to improve the way development data are produced, managed, and used. They bring together diverse teams of collaborators from around the world, and are focused on solving challenges in low and lower middle-income countries in Sub-Saharan Africa, East Asia, Latin America, and South Asia.

Following the success of the first round of funding in 2016, in August 2017 we announced a $2.5M fund to support Collaborative Data Innovations for Sustainable Development. The World Bank’s Development Data group, together with the Global Partnership for Sustainable Development Data, called for ideas to improve the production, management, and use of data in the two thematic areas of “Leave No One Behind” and the environment. To ensure funding went to projects that solved real people’s problems, and built solutions that were context-specific and relevant to its audience, applicants were required to include the user, in most cases a government or public entity, in the project team. We were also looking for projects that have the potential to generate learning and knowledge that can be shared, adapted, and reused in other settings.

From predicting the movements of internally displaced populations in Somalia to speeding up post-disaster damage assessments in Nepal; and from detecting the armyworm invasive species in Malawi to supporting older people in Kenya and India to map and advocate for the better availability of public services; the 12 selected projects summarized below show how new partnerships, new methods, and new data sources can be integrated to really “put data to work” for development.

This initiative is supported by the World Bank’s Trust Fund for Statistical Capacity Building (TFSCB) with financing from the United Kingdom’s Department for International Development (DFID), the Government of Korea and the Department of Foreign Affairs and Trade of Ireland.

2018 Innovation Fund Recipients

Data science competition: predicting poverty is hard - can you do it better?

Tariq Khokhar's picture
 

If you want to reduce poverty, you have to be able to identify the poor. But measuring poverty is difficult and expensive, as it requires the collection of detailed data on household consumption or income. We just launched a competition together with data science platform Driven Data, to help us see how well we can predict a household’s poverty status based on easy-to-collect information and using machine learning algorithms.

The competition supplies a set of training data with anonymized qualitative variables from household surveys in 3 countries, including the “poor” or “not poor” classification for each observation.

The challenge is to build models which can accurately classify households from a different set of test data (with the poor/not poor classification removed!) for the same 3 countries, and then submit them for scoring. Performance is measured by the mean log loss for the 3 countries which tells us how accurate the classification models developed are.

Prizes are $6,000; $4,000; and $2,500 for the top 3 performing entries, plus a $2,500 bonus prize for the top-performing entry from a low- or lower-middle income country. The deadline for entries is February 28th 2018.

You can read the full problem description and enter the competition here, and see the Driven Data team’s “benchmark solution” based on a random forest classifier.

Good luck - we look forward to seeing your solutions!

Is your country LGBTI inclusive? With better data, we’ll know

Clifton Cortez's picture

The World Bank is developing a global standard for measuring countries’ inclusion of LGBTI individuals.

They laughed in our faces … but then we showed them the data

By the early 1990s, Dr. Mary Ellsberg had spent years working with women’s health in Nicaragua. Armed with anecdotes of violence against women, she joined a local women’s organization to advance a bill criminalizing domestic violence.

When presented with the bill, lawmakers “pretty much laughed in our faces,” she explained in a 2015 TEDx talk. “They said no one would pay attention to this issue unless we got some ‘hard numbers’ to show that domestic violence was a problem.”

Dr. Ellsberg went back to school and wrote her doctoral dissertation on violence against women. Her study showed that 52% of Nicaraguan women had experienced physical or sexual abuse by an intimate partner. Subsequently, the Nicaraguan parliament unanimously passed the domestic violence bill.

Later, the World Health Organization used Dr. Ellsberg’s indicators to measure violence against women in countries across the world, which showed the global magnitude of the problem.

“One out of three women will experience physical or sexual abuse by her partner,” Dr. Ellsberg said. Because of the data, “violence against women is at the very top of the human rights agenda.”

Dr. Ellsberg knew that domestic violence was a problem, but it was data that prompted leaders to combat the issue.

Similarly, there are plenty of documented cases of discrimination and abuse against lesbian, gay, bisexual, transgender, and intersex (LGBTI) people. But what’s the magnitude of the discrimination?

Pages