Exploring Features, Customer Emotions, and Evaluations of Airbnb Online Experiences: A Framework for Virtual Tourism Experiences

Abstract

This study proposes a conceptual framework grounded in the Stimulus-Organism-Response theory (S-O-R) to investigate the relationship between virtual tourism experience features, customer emotions, and their evaluations. Through a hybrid approach combining automated aspect extraction and manual annotation, five key features of online experiences (OEs) are identified: host attributes, activity design and experiential value, technology and technical tools, value for money, and personalization. Leveraging deep learning-based emotion analysis, the research maps customer emotional states along a spectrum from positive (e.g., admiration, gratitude, approval, optimism) to neutral or negative (e.g., disappointment, disapproval). Empirical findings reveal that OE features systematically trigger emotional responses, as validated through sentiment classification techniques. Moreover, ridge regression analysis confirms that these emotional states are strong predictors of OE ratings, explaining a significant proportion of variance in customer evaluations. This work advances the theoretical understanding of virtual tourism experience design and offers actionable strategies for platform optimization.