Social Media-Based Emergency Management to Organize Civilian Volunteers

In today's world, man-made and natural disasters happen almost every day such as fires, earthquakes, floods, and terrorist attacks. Many organizations are in charge to handle such situations and minimize the casualties. States have institutions, equipment, and officials available for this purpose. The procedure to be followed and the plan to be implemented in the event of a disaster are determined. Since resources are not unlimited, these procedures must offer the most optimal solutions in terms of cost. However, since time is a very important parameter to save lives in the event of a disaster, the system must be able to produce results very quickly. These issues are considered in the functioning of official institutions. But apart from that, there are benefits that civilians can also provide. This support, if not organized and managed, can lead to chaos rather than benefit. Information on the needs of individuals and the support they can offer can be collected on social media. Once the information is collected, an optimization procedure can be applied for coordination. In this study, considering that the information also includes physical coordinates, the situation is modeled as a clustering problem Cost is calculated in terms of travel and initial facility setup costs, and K-Means and Mean-Shift have been applied. Test results show that MeanShift algorithm reveals better optimization.