Social media can save lives when disaster strikes

Twitter, Facebook and other social media have created new opportunities to inform and respond when disaster strikes. In her PhD project at the IT University of Copenhagen, Alivelu Mukkamala looks into how data from social media can be used to improve disaster response.

Researchsocial media

Written 16 February, 2016 09:00 by Vibeke Arildsen

A few minutes after a major earthquake hit Nepal in April 2015, a mountaineer on Mount Everest sent out a call for help on Twitter: "Basecamp has been severely damaged. Our team is caught in camp 1." During the 2015 Chennai floods, rescuers and affected people used social media to plead for humanitarian aid and drinking water. As the Paris terrorist attacks were unfolding in November, the media showed eyewitness videos, while Parisians offered each other shelter on Twitter.

There are countless examples of how social media has turned the information chain upside down when disasters and terrorist attacks strike.

"Previously, the authorities and the media provided information to the population, but now ordinary citizens can share valuable information through social media, thereby contributing to creating an overview of the extent and character of the disaster," says Alivelu Mukkamala, a PhD student at the IT University of Copenhagen.

In her PhD project, she investigates how data from Twitter, Facebook and other social media can be used to make disaster relief activities more efficient. The challenge is extracting the relevant bits of information from the enormous amount of updates flooding social media in the wake of disasters.

"There were over 15 million tweets about Hurricane Sandy when it hit the US East Coast in 2012. A lot of it was noise - for instance people discussing the hurricane in Europe, and people talking about leaving work early because of the hurricane. We need a tool that makes it possible to extract useful information from the vast amount of user-generated information on social media in connection with such disasters."

Time is crucial

In disaster situations, time is the most crucial factor, and it takes many days to find the useful information on social media manually. With the method we are developing, data analysts would be able to extract the information about urgent needs quickly.

Alivelu Mukkamala, PhD student

The goal of Alivelu Mukkamala’s research is to develop a method that can help authorities, aid organizations and rescue workers to use social media to assess the situation when disaster hits. Such a method would make it easier to determine where the need for help is the most urgent.

The first part of the project entails compiling a list of hundreds of words and phrases that indicate needs and urgencies of people.

"In disaster situations, time is the most crucial factor, and it takes many days to find the useful information on social media manually. With the method we are developing, data analysts would be able to extract the information about urgent needs quickly."

In the next phase of the project, the method will be tested on data from previous disasters. When finished, the goal is to integrate the method into the existing open-source tools such as geo-location trackers, making it possible to create accurate assessment of when and where to distribute aid. The Danish Emergency Management Agency has already taken an interest in the project and there are plans for a collaboration later in the process. 

New types of collaborations

According Alivelu Mukkamala, not only authorities and organizations can benefit from social media during disasters. Ordinary people, too, have found new opportunities for cooperation.

"Through the social media, people can both ask for help, offer help and share information. Usually, people are very careful about sharing personal information online, but during emergencies, people will share phone numbers, offer accommodation to strangers and warn strangers about staying away from certain areas. This type of collaboration between people was not really possible before social media, and I think that's really amazing.”

Alivelu Mukkamala’s PhD project runs from March 2015 to February 2018.