Uncovering Spatiotemporal Characteristics of Human Online Behaviors during Extreme Events - Supplementary Information

In response to an extreme event, individuals on social media demonstrate interesting behaviors in information seeking and sharing, depending on their backgrounds. Existing studies have attempted to analyze and understand the regularities of human responses during an extreme event, e.g., their spatiotemporal patterns. However, most of them focus on the general patterns of human collective online activities from a population perspective. In reality, individuals with different backgrounds, e.g., living in different geographic locations, will behave differently. Therefore, it remains to be a crucial issue for us to quantitatively understand human collective online behaviors during extreme events, especially from an individual perspective. Due to the difficulty in measuring individual behaviors and the lack of datasets from which the underlying dynamical processes may be inferred, it is more challenging for us to probe into the dynamics of human collective online behaviors. By making use of the large-scale datasets of posts and search queries collected from Twitter and GoogleTrends during various events, it is now possible for us to analyze individuals’ online behaviors and then reveal distinct characteristics (i.e., heterogeneous spatiotemporal patterns) of collective online responses to extreme events.