Journal on Multimodal User Interfaces Обработка сигналов
Enhancing affective experience with an Artificial Commensal Companion through personality-based conversational adaptation
Аннотация
Eating is a fundamentally social activity, yet many people increasingly dine alone, often accompanied by technology. Artificial Commensal Companions (ACCs) have been proposed to support the social dimension of meals, but empirical evidence of their impact and the role of personalization remains limited. This paper examines users’ affective responses to interacting with an ACC during meals, focusing on personality-based conversational adaptation. A mixed-methods study was conducted in a naturalistic lunchtime setting. Participants interacted either with a personalized ACC, whose conversational style was adapted to the user’s personality, or with a non-personalized version of the same agent. Quantitative results show a significant reduction in Negative Affect following the interaction, while Positive Affect remained stable. This reduction was larger in the personalized condition, suggesting a potential benefit of personality-based conversational adaptation. The qualitative analysis shows that ACCs can provide a valuable and interesting experience, especially for individuals eating alone, and may lead users to appreciate human relationships even more. Overall, the results suggest that ACCs may support affective well-being during meals and highlight the importance of tailoring conversations.