This paper proposes a system that automatically infers passengers' emotional states based on flight delay situations and recommends music tailored to each emotional state. The system collects flight schedule information through the Google Calendar API, retrieves real-time delay and weather data through aviation and weather APIs, and classifies emotional states using a rule-based approach grounded in Russell's (1980) Circumplex Model of Affect. Emotional states are determined by combining three factors: delay duration, weather conditions at the departure airport, and scheduled departure time. A total of 12 representative emotional states are defined, each mapped to specific music audio features (valence, energy, tempo, danceability, and instrumentalness) defined by the Spotify Web API. The recommendation engine adopts a content-based filtering approach to match appropriate tracks to each emotional state. A simulation experiment with sample flight scenarios confirmed the system's operational feasibility. This work contributes to the emerging field of context-aware emotion regulation by presenting an automated, theory-grounded framework for entertainment curation during stressful travel situations.
jaehan kim (Tue,) studied this question.