Exploring situational factors influencing tourists’ evacuation behavior

Major disaster events pose severe threats to tourists in hazard-prone areas. Although prior studies explored factors influencing individuals’ protective actions, they oversimplified the complexity of making evacuation decisions and evacuation behavior. In reality, evacuation decision reflects a chained series of choices. The purpose of the study is to explore situational factors influencing tourists’ evacuation behavior during a major hurricane. Methodologically, a machine learning approach is used to clean social media data, predict evacuation behavior, and analyze the main topics of online tourists’ discussion during disaster events. The results show proactive tourists’ behavior towards receiving and responding to information through media channels and a high valuation of stakeholders’ capability of planning the emergency. Interestingly, both evacuating and sheltering-in-place tourists tend to be concerned about similar topics regardless. The implications are discussed.