Conflict remains one of the most pernicious ills facing the world today. Destruction, instability and related poverty and remain widespread; as the World Bank’s April 2011 report notes, conflict “has become a primary development challenge of our time. One-and-a-half billion people live in areas affected by fragility, conflict, or large-scale, organized criminal violence, and no low-income fragile or conflict-affected country has yet achieved a single United Nations Millennium Development Goal.” At the time of writing, bombings and indiscriminate killings continue in Iraq due to rising Sunni-Shia sectarian violence; Afghanistan is facing rising violence and hostility due to a drawdown of coalition forces & contested national elections; devastation continues in Syria’s civil war. (A list of ongoing global conflicts can be found here).

We have seen big data used to combat other global problems; it can optimise and mobilise disaster relief, predict and curb disease outbreaks such as flu and dengue fever, and help governments predict and prepare for wildfires. But can big data prevent conflict?

If there was ever a time that data could be leveraged as a conflict prevention tool, it’s today. As FP highlights, the primary reason is we have more data than ever before; the statistic that we created more data in 2012 than in the rest of human history combined is oft-cited. In the past decade, internet access in Middle East and Africa has grown 2634% and 3607% respectively. Not only do we have more data, but we also have more sophisticated ways of using it; new machine learning algorithms, such as those employed by the International Peace Institute, have offered “substantial improvements in accuracy and performance” in conflict prediction tools. Finally, it’s not just the volume of data that has shifted; it’s the variety. Social media and user-generated content, coupled with the emerging field of sentiment analysis, means we get insights into what’s going on on a human level- but also crucially how people are feeling and responding to events.

So how are the data and technologies actually being applied? Many organisations are gathering social media, satellite information, news-scraping apps, blogs and more to get a better picture of where conflicts are occurring, or likely to occur. The list of such programmes is extensive- the US Defense Department have their Information Volume and Velocity Programme; the UN have Global Pulse; the United States Agency for International Development have a Foreign Assisstance Dashboard; the CIA have an Open Source Indicators Programme.

Clearly, NGOs and international organisations alike believe there’s value in such data-crunching initiatives. But measurable results are sparse; few have documented success cases. Of course, quantifying absence of violence is challenging, but many have highlighted a void between data gathering & analysis, and actually transforming forecasts into actionable responses. Africa’s Conflict Early Warning and Response Mechanism (CEWARN)has been widely acknowledged as somewhat of failure based on such grounds; data capture was successful, but actually turning this into a viable prevention strategy was not.

There are, however, some success stories- but more often than not, these come from smaller, localised organisations rather than global initiatives with a more remote understanding. One such example came during the 2013 Kenyan presidential elections. The previous elections in 2010 led to wide-spread violence, leaving 1,300 dead and 600,000 homeless. A web of NGOs collaborated on early-warning and early-response technologies to ensure the devastation of 2010 wasn’t repeated. One nonprofit, Ushahidi (“witness” in Swahili) developed open source data collection and mapping software, which was then used by the Umati project to monitor social media, reports and blog posts to detect rising tensions and hate speech. Mobile phone provider Safaricom donated 50 million text messages to the Sisi Ni Amani (“we are peace”) project, so that it could respond to hate speech and calls to violence. Texts were used widely in the prior conflicts to incite hatred and organise violence; Si Ni Amani wanted to harness the power of texts to diffuse tensions and spread co-operation.

The success of their campaign is detailed on their blog. One particularly striking example comes from Dandora, one of the areas which experienced “the most tensions and incidents in the days leading up to and following the elections”. The blog details:

On election day, youths supporting a particular political party overwhelmed police in Phase 4 to prevent voting. This was brought under control by backup security, and SNA-K sent a message.
Feedback from residents included comments that, the message “helped to calm down the situation” and “was sent at the right time,” because, according to partners “they think the whole world was watching Dandora. Everyone knew what was happening.”

Following a second message, our Dandora co-ordinator summarized feedback from eight outreach workers in the area. They said: “the message helped to maintain calm, reminds us of our community, makes us be united, shows someone thinks about Dandora, reminds us to be peaceful all the times, and thanks for reacting and responding to our concerns.”

In the Burnt Forest and Kariobangi North area, another co-ordinator stated “The message had a real impact. People stopped and were looking at their phones. They were gathered in groups and talking about politics”.

This type of approach aligns well with recent strategies advocated by McKinsey. They outline a “Designing for Development” approach to small wars, which involves a blend of remote observation (the kind of big data analytics being harvested by the UN, CIA et al above) with “contextual understanding”, on-the-ground information collated by people from within the instable communities.

They highlight that big data has a role to play, but that communities and people need to be at the heart of conflict prevention strategies, as they are the “engines of social transition”. One of the key challenges they highlight is putting the right creative problem solvers within these communities to prevent conflict. Such people face a host of challenges- building credibility and trust, mitigating risk for themselves and those they work with, representing interests between different communities and groups in an unbiased manner. Yet, as McKinsey points out, facing such challenges is essential- the only people with sufficient understanding of the interests at stake are within the communities, not thousands of miles away watching unrest grow on a data-driven dashboard.

There is of course another, more sinister face to big data in conflict- big data that’s used to win wars, not prevent them. Data has a long and bloody history in helping to win wars. A recent article detailed how British Intelligence amassed operational records and data garnered from accompanying missions to optimise attacks again German U-Boats in World War II. They used the data to discover the optimal setting for depth charges (25 feet) and the most effective bombing patterns. By the end of the war, 784 of the 830 U-boats deployed were destroyed, a staggering 94%. Only this week, the US Navy issued a call for papers on how they best leverage their big data. What we unfortunately see in this bloody, brutal realm of big data is what is fundamentally lacking in the field of conflict prevention- a direct path being forged between gaining insights and doing something about them.

Let’s hope that in the future, a clearer link between the data scientists predicting conflict outbreaks and the resolution specialists who can prevent them is established.

(Image credit: Flickr)

Eileen McNulty-Holmes – Editor


Eileen has five years’ experience in journalism and editing for a range of online publications. She has a degree in English Literature from the University of Exeter, and is particularly interested in big data’s application in humanities. She is a native of Shropshire, United Kingdom.


Interested in more content like this? Sign up to our newsletter, and you wont miss a thing!



Previous post

Wimbledon & Big Data: Predictive or Missing the Mark?

Next post

Infographic: Where to Find & Access Big Data