Psychologically And Physiologically Measuring Music-induced Emotion And Its Application In Service Design

This present study explored music-induced emotion through psychological andphysiological measurement. This study focused on how the following factors impactuser’s emotion: a) mode (major, minor and pentatonic); b) rhythm pattern (firm andflowing); c) MV(music video) type (narrative, live, parody). Then an EEG-based musicinteraction system—MusiEEG—was designed and developed by integrating musicaltheories and emotion measurement into musical service design.To this end, three studies were conducted. The first study focused on howto measure emotion induced by pop music psychologically and physiologicall,and how mode and rhythm pattern would impact emotion. An experiment wasconducted and12participants were recruited. The second study was conductedto study how MV (Music Video) type and rhythm pattern would influenceinduced emotion.30participants took part in the experiment. The third studydesigned and developed an EEG-based music interactionsystem—MusiEEG—based on the theoretical and methodological findings.MusiEEG was able to predict users’ music preference and personality traits bymeasuring their EEG data during music, and help them to find others who are perfectly“matched” with them according to the law of harmony.Musically-related results indicated that1) mode influences arousal, valencesignificantly. More specifically, there are significant difference betweensublimity and HRV between pentatonic and western modes;2) rhythm patterninfluences arousal and sublimity significantly;3) MV type influences arousal,sublimity, and vitality significantly;4) Effects of MV type and rhythm patternare interdependent. Rhythm pattern impacts sublimity induced by narrative andlive MV, while it has no effects on sublimity within parody MV. On the otherhand, vitality differs significantly in terms of rhythm pattern when the MV isnarrative and parodical, while the live performance MV seems not beaffected.As for EEG-based emotion measurements, EEG power spectral analysisserves sound criteria for emotion recognition and music classification. The thetaasymmetry in left and right hemispheres is significant when people are listening to pop music; beta power is correlated with valence, vitality, tension andsublimity, and differs significantly in terms of MV type. gamma power iscorrelated with arousal, and has distinctive patterns within music in major mode,and gamma power in different types of MV has siginifcant difference. alphapower is correlated with arousal level.Theoretical and methodological findings of this study provide a solidfundation for the design and development of EEG-based music interactionsystem. Design strategies were provided based on the findings. Design of musicinteraction service was also explored with the implications.