Big data's big role in humanitarian aid

Hundreds of thousands of refugees streamed into Europe in 2015 from Syria and other Middle Eastern countries. Some estimates put the number at nearly a million.

The sheer volume of people overwhelmed European officials, who not only had to handle the volatile politics stemming from the crisis, but also had to find food, shelter and other necessities for the migrants.

Sweden, like many of its European Union counterparts, was taking in refugees. The Swedish Migration Board, which usually sees 2,500 asylum seekers in an average month, was accepting 10,000 per week.

"As you can imagine, with that number, it requires a lot of buses, food, registration capabilities to start processing all the cases and to accommodate all of those people," says Andres Delgado, head of operational control, coordination and analysis at the Swedish Migration Board.

Despite the dramatic spike in refugees coming into the country, the migration agency managed the intake — hiring extra staff, starting the process of procuring housing early, getting supplies ready. Delgado credits a good part of that success to his agency's use of big data and analytics that let him predict, with a high degree of accuracy, what was heading his way.

"Without having that capability, or looking at the tool every day, to assess every need, this would have crushed us. We wouldn't have survived this," Delgado says. "It would have been chaos, actually — nothing short of that."

The Swedish Migration Board has been using big data and analytics for several years, as it seeks to gain visibility into immigration trends and what those trends will mean for the country. Delgado says its Qlik Technologies analytics tools gave his agency, and Sweden, the ability to prepare for the rush of refugees in a way it couldn't have years ago.

"We [once] looked into the future with our back toward the horizon. That wasn't really a good form of control of the operations," Delgado says. "We wanted to turn around 180 degrees and look into the future. We got a tool that would give us the capability to gather information [and] process all this data and look into predictions of the upcoming year so we could be three or six months ahead in our planning."

The Swedish Migration Board's success in anticipating, and thus being able to prepare for, the refugee crisis of 2015 is impressive enough. But that's only the start of the story. The real story lies in what this agency and other humanitarian organizations are starting to do with big data and analytics. Like their corporate brethren, humanitarian and peace-building organizations are harnessing data to gain insight into the people and problems they're tasked with handling. And just like the for-profits, they're finding that their efforts are paying off.

"Can big data give us peace I think the short answer is we're starting to explore that. We're at the very early stages, where there are shining examples of little things here and there. But we're on that road," says Kalev H. Leetaru, creator of the GDELT Project, or the Global Database of Events, Language and Tone, which describes itself as a comprehensive "database of human society."

The topic is gaining traction. A 2013 report, "New Technology and the Prevention of Violence and Conflict," from the International Peace Institute, highlights uses of telecommunications technology, including data, in several crisis situations around the world. The report emphasizes the potential these technologies hold in helping to ease tensions and address problems.

The report's conclusion offers this idea: "Big data can be used to identify patterns and signatures associated with conflict — and those associated with peace — presenting huge opportunities for better-informed efforts to prevent violence and conflict."

That's welcome news to Noel Dickover. He's the director of PeaceTech Data Networks at the PeaceTech Lab, which was created by the U.S. Institute of Peace (USIP) to advance USIP's work on how technology, media and data help reduce violent conflict around the world.

Such work is still in the nascent stages, Dickover says, but people are excited about its potential. "We have unprecedented amounts of data on human sentiment, and we know there's value there," he says. "The question is how to connect it."

Dickover is working on ways to do just that. One example is the Open Situation Room Exchange (OSRx) project, which aims to "empower greater collective impact in preventing or mitigating serious violent conflicts in particular arenas through collaboration and data-sharing."

The OSRx is a platform being developed to help peace-builders working on the ground, creating a framework for them to think about how data can be used effectively in their efforts to end conflicts in the short term and in the future.

According to the PeaceTech Lab's website, the OSRx will "provide the ability to find, collect, analyze, visualize and publish conflict related data" as well as connect local peace-builders with technology and technologists.

Dickover says the exchange is set to launch early this year with three dashboards available in 360 countries, providing information from social media as well as structured indexes, such as data sets on economic fragility, in areas around the world.

"The idea is to take the real-time information and almost forecast what can happen," he says. "And what's really impressive is we're going to make it publicly accessible. We're really looking to make this accessible for local peace-builders to use for their own campaigns and the work they're doing."

He points to a project in Myanmar (formerly Burma) as an example of how it can be used. The project has civic organizations working together to monitor hate speech with the goal of taking countermeasures before belligerent discourse spins out of control into actual violence. The organizations learned how to use technologies, including big data and analytics tools, to scout for and analyze incidents of inflammatory speech so they could mobilize teams to take action — such as flooding social media with more positive messages.

This involves much more than simply watching for offensive Facebook posts. The project takes in reams of data, mostly from locally generated and publicly accessible online social media sources, and feeds it into an artificial intelligence engine. This process will help the system develop and fine-tune the algorithms that will analyze data.

"It will create the monitoring tool that [people] can then use to take action," Dickover says.

The ability to do what Dickover describes is far from wishful thinking. In fact, that goal mirrors what corporations and other types of organizations around the world are already doing with big data. Companies, healthcare agencies and others aren't just compiling information from numerous sources to understand what has happened in the past. They're using it to grasp what's happening right now, in real time, and — more importantly — what will likely happen in the future and how they can take advantage of future developments.

For example, leading-edge companies generally use big data and analytics to drive sales and lower costs by, for example, figuring out what promotions will get specific customers to buy their products.

Forward-thinking peace-builders and humanitarian organizations are beginning to talk in similar terms, discussing — at least in theory — how they can use information to determine likely future outcomes and then develop plans to affect those outcomes.

"Can we use predictive and prescriptive analytics That's the ultimate goal," Dickover says.

But he and others say that's a long way off. There are, to be sure, huge obstacles to achieving such noble objectives. Members of this community struggle with collecting and organizing the right data in the right format for emerging analytics initiatives. They also have a hard time getting required resources, whether it's identifying the best technologies for their needs or finding the money to pay for the technology and talent that can help them achieve their goals. And they strain to understand how to best use the data and analysis they can get, as they weigh whether empowering local efforts is more effective than arming high-ranking policymakers.

Those challenges are parallel to the ones that for-profit organizations face, Leetaru says. But just as those challenges aren't deterring corporations from pushing forward with their analytics programs, he says, they shouldn't stop peace-building and humanitarian groups either.

"I see interest all around the world in leveraging this technology. There's phenomenal interest," Leetaru says.

Still, he and others are realistic. This community generally has fewer resources — money and personnel — than businesses do, and that makes it harder to push forward. Additionally, the data that nonprofit organizations need — data gathered from and about individuals and groups in developing countries where conflicts tend to be more likely — isn't being generated or collected at anywhere near the same rate as commercial data is in the Western world.

Leetaru is trying to change that. GDELT monitors print, broadcast and Web news media around the world, in more than 100 languages, every moment of every day. Its collection spans from Jan. 1, 1979, to the present day, with daily updates. And it's available for organizations and individuals to use to gain insight into the circumstances that shape events.

"If we can create a live catalog of everything on Earth, that increases awareness — it gives voice to people around the world," Leetaru says.

For now, data-driven projects tend to be more targeted.

Take, for example, some of the initiatives run by the U.S. Agency for International Development (USAID).

The agency is analyzing geospatial data, satellite imagery and other information to identify regions of the world where unsustainable uses of water (such as fish farming) are common, so it can advocate for changes that will head off droughts, says Craig Jolley, an AAAS science and technology policy fellow at USAID.

He also cites USAID's work with Pakistani power companies to install smart meters and then analyze usage patterns to help better manage electric capacity and avoid energy shortages and blackouts that can feed into civil unrest.

Eric King, a specialist on humanitarian work and IT at USAID, says that as more data is collected from around the world, the potential to run additional projects such as these becomes possible. But, like others in the field, he stresses that the work is in its infancy.

He also points out that while big data and analytics, like all other technologies, can help foster change, they have their limits. It's best to be realistic in that regard — and remember that they're just tools to help people be more effective.

"Technology is far from a panacea," King says. "People will be the ones to solve these problems, but the technology will help them do it. Technology is really helping the people who are working on these problems on the ground do their job better."



Mary K. Pratt

Zur Startseite