Assessing emotions in online stories: comparing self-report and text-based approaches

Abstract

The core product of tourism is experience and therefore is inherently emotional in nature. As such, online storytelling has become an essential destination promotion strategy as it effectively conveys information while at the same time builds emotional connections with potential travelers by encouraging imagination and involvement. Understanding and measuring travelers’ emotional responses to online stories are essential to developing effective advertising strategies. Traditionally, the self-report approach has been employed whereby informants are asked to identify their emotions. With the development of advanced machine learning techniques, new tools such as sentiment analysis have been developed to detect the emotions conveyed by text. In this study, we summarized the results of a comparative analysis of emotion detection using both self-report method and sentiment analysis. The results show general consistency between the two approaches in terms of overall descriptive statistics as well as the relationships between emotions and various components of advertising response.