Spatial and temporal analysis of human movements and applications for disaster response management: Using cell phone data

This survey study examines cell phone usage data and focuses on the application of the data to disaster response management. Through the course of this study, the structure of cell phone usage data and its characteristics will be reviewed. Cell phone usage data provides us with valuable information about human movements and their activities. The uniqueness of the data is that it contains both spatial and temporal information and this information is free of fixed routes such as roads or any preset data capturing timing. In short, it is a very fluid kind of data which reflects our activities as humans with freedom of movement. Depending on data extraction methods, the data server can provide additional information such as application activities, battery level and charge activities. However, cell phone usage data contains shortcomings including data inconsistency and sparseness. Both the richness and the shortcomings of the data expose the hurdles required in data processing and force us to devise new ways to analyze this kind of data. Once the data has been properly analyzed, the findings can be applied to our real life problems including disaster response. By understanding human movement patterns using cell phone usage data, we will be able to allocate limited emergency resources more adequately. Even more, when disaster victims lose their cell phone functionality during a disaster, we might be able to identify or predict the locations of victims or evacuees and supply them with necessary assistance. The results of this study provide some insights to cell phone usage data and human movement patterns including the concentration of cell phone activities in specific zones and rather universal cell phone charging patterns. The potential of the data as a movement analysis resource and the application to disaster response is apparent. As a base to leverage the study to the next level, a possible conceptual model of human movement factors and data processing methods will be presented