Emotions engines and emoji magic: Unlocking discrete emotional drivers and emoji moderation in destination public opinion through text mining

Abstract

Rapid dissemination of information on social media challenges effective destination image management. Through text mining, this study explores how public emotions triggered by the tourism crisis event on social media affect users’ engagement behavior. It is found that specific discrete emotions—particularly anger and love—significantly enhance user engagement, identifying them as the most influential negative and positive drivers, respectively. Unlike previous studies that merely consider the presence or absence of emojis, this research shows that a higher proportion of emoji use weakens the impact of these emotional drivers. The study further uncovers the internal structure and the dynamic evolution of high-impact emotions using structural topic modeling. These findings provide actionable insights for tourism stakeholders, enabling them to refine public opinion strategies and respond swiftly to online emotion shifts.