In a collaborative effort between Jose Marquez of Stevens and Gabriela Gongora of Carnegie Mellon, new research published in the Journal of Risk Analysis shows how social media can be used to measure social cohesion during a natural disaster. This research acts as a stepping stone for further investigations into the field, showing how it may be possible for community leaders to use social cohesion to improve their disaster mitigation policies.
Social cohesion, as defined by Marquez and Gongora, “is believed to be the glue that holds society together,” and “can impact how people recover psychologically from hurricanes and how willing they are to collaborate in rebuilding efforts.” The researchers suggest that by studying this property, they can learn more about the networks that are created between users on social media platforms and gain incredible insight into how information spreads during a crisis. By understanding this, government agencies and crisis management teams could accurately and efficiently spread important information regarding their current situation. They might also be able to pinpoint certain areas of concern, and spread recovery efforts more effectively.
In their investigation, Marquez and Gongora analyzed the data surrounding three major concurrent hurricanes from 2017: Hurricane Harvey in Texas, Hurricane Irma in Florida, and Hurricane Maria in Puerto Rico. First, data was collected from Twitter and then checked to see if it corresponded with one of the occurring hurricanes. Their accounts were then categorized based on their purpose, “citizens” if they weren’t verified, or into 1 of 8 groups if they were, such as “NGO,” “Media,” or “Government.” The accounts were then assembled into a network of connections, based on sent messages and retweets. From this, researchers were able to analyze social cohesion.
Marquez and Gongora found that the number of tweets they measured through Twitter can provide good insight into the status of a natural disaster. With Hurricane Harvey, there was a slow amount of buildup in social media usage before the storm, which peaked when the storm made landfall. After the storm there was still a good amount of social media usage, indicating to the researchers that recovery efforts were underway. The same is true for Hurricane Irma, a buildup of activity before the storm, peaking at the same time the storm does, and then slowly falling off as recovery efforts evolve. Hurricane Maria emulates the same effects as the other two, but over a longer period of time. There is a longer buildup period due to speculations weeks before landfall, and a longer recovery period because of the extended recovery period for the hurricane.
The social cohesion measured can also provide similar insights. While not peaking with the storm, the social cohesion still extends well before and after, showing how communities will strengthen in times of crisis. It also shows how as recovery efforts progress and then finish, social cohesion diminishes, indicating the end of recovery efforts.
This research shows the potential benefits that social media can provide for relief efforts. Analyzing social media can help governments understand how a community has been impacted by a disaster, and how to mitigate these effects and execute a proper recovery effort. Although this study focuses mostly on the potential ways to identify social cohesion, it opens the door to future research that can further understand this phenomenon, and apply it to real-world scenarios in real-time.
Be First to Comment