ECG-Based AI Emotion Recognition for User Journey Mapping in UX Evaluation

Traditional user experience (UX) evaluations rely heavily on post-task recall and subjective questionnaires, which overlook moment-to-moment emotional changes during interaction. Existing approaches using physiological or AI-driven emotion recognition rarely link these signals to structured UX analysis or concrete design decision-making. This study proposes an AI-assisted framework that uses ECG-based emotion recognition to map users’ emotional states along the user journey and integrates these emotion traces into a structured UX evaluation process supported by Kano-based requirement analysis. In a case study comparing two mobile coloring applications (n = 30), the framework captured emotional fluctuations not reflected in retrospective reports and revealed 41% more UX pain points than recall-based journey maps. These findings indicate that combining physiological signals, AI models, and journey mapping can enhance the precision of UX assessment and support more evidence-based design decisions, although validation is currently limited to a small, lab-based study.