Spatial Network Activity Summarization for Disaster Response

Research overview and Potential Significance: My research explores novel computational techniques for enhancing situational awareness for disaster response, which is a national imperative [12]. Beyond the unquantifiable costs of injury and loss of life from disasters, economic damages from Hurricane Sandy alone in the United States exceeded $60 billion [13]. Disaster response during emergency management includes action taken immediately after a disastrous event with the aim of saving life, protecting property, and dealing with immediate disruption, damage, or other effects caused by the disaster (e.g., hurricanes, earthquakes, terrorist attacks, etc.) [1]. The 2010 Haiti earthquake saw numerous requests for relief including requests for food, water, and medicine. Emergency managers are tasked with making resource allocation decisions based on the locations of the affected population. Figure 1(a) shows the locations of a subset of various relief requests in Port-au-Prince, Haiti from January to March 2010. An important aspect of disaster response is to understand where people are in a given area (e.g., situational awareness). In other words it is important to be able to summarize people’s location on a map and place them into different groups depending on their location. In order to make informed, critical resource allocation decisions given such location information, emergency managers need tools to assist them with situational awareness, which I conceptualize in my work as the spatial network activity summarization problem (SNAS).