Assessing uncertainty in VGI for emergency response

This research project examines the geographic data produced by volunteers via the Ushahidi web platform in response to the earthquake that struck Haiti in January 2010. Volunteers translated messages (text, e-mail, and voice) submitted by victims of the earthquake, categorized each message into a primary ‘emergency need’ category and subcategory, and georeferenced each message on a dynamic web-based map. Initial inspection of the categorized data indicated discrepancies between the emergency need submitted by victims and the subsequent categorization of the emergency need. Analysis of the main categorical data illustrated that 50% of the messages were mis-categorized by the volunteers, failing to convey the main idea of the victim’s message. At the subcategory level, approximately 73% of the messages failed to convey the main idea of the messages. These numbers are higher than the estimate of 36% error in categorization produced in an independent review of the Haiti Ushahidi database. While the volunteer response to the Haitian earthquake represents a paradigm shift in emergency response and victim empowerment that has been repeated in numerous natural and man-made disasters around the world, this study suggests the need for more research on the quality of the categorization (i.e., attribute data) of volunteered emergency data.