UC Santa Barbara

The general aim of this research is to understand how social media content, generated by the general public via Twitter, can be used with volunteered geographic information and official data sources to provide information about how weather-related disaster events are experienced at a local scale. Two main objectives are addressed in this study: (1) understanding how spatio-temporal information embedded in tweets can be leveraged to summarize their content and identify local effects and damage caused by weather-related disaster events; and (2) exploring how both geographic information related to hazards and validated citizen reports can be used for assessing the reliability of tweets and to prioritize those most useful for emergency management. This research contributes to the understanding of how social media data contributed by the general public can be used to inform emergency response. Additionally, it shows that different spatial clustering methods and spatial aggregation can work both to enhance content analysis in order to summarize and identify tweets containing actionable information, and for conflating tweets with authoritative geographic data to assess tweets reliability.