A dataset from ten disasters for studying location descriptions and training AI models

Location information is highly important during a natural disaster for reaching victims and coordinating response and relief efforts. In recent years, people increasingly use social media to post messages during a disaster, and describe the locations of victims, damages, difficult situations, and relief resources. Many of these location descriptions are in the forms of detailed and multi-entity descriptions, such as door number addresses, road intersections, and highway exits. Currently, there is limited availability of datasets containing these detailed location descriptions, which hinders their understanding and automatic extraction. This paper fills this gap by providing a dataset that contains labeled location descriptions from ten disasters in the United States and in five disaster types: hurricanes, floods, wildfires, tornados, and winter storms. The messages containing location descriptions are collected from the social media platform Twitter/X, and we describe the collection, labeling, and validation of this dataset. This dataset can be used for studying the ways people describe locations under disaster contexts and for training AI models to extract these important locations.