M-Motion

Abstract

Studies show that music is an effective way of emotional induction and can change the emotional behavior of the user. The current methods of recommendation and reproduction of digital music, basead on emotion, require manual user interaction and most of them select from a set of generic music and not from the user's taste. Therefore, this paper proposes a mobile application, m-Motion, that elects a subset of songs from the user, considering his current emotional state and supports the user to achieve a desired emotional state. An experiment was conducted by a group of 8 users. We collected the current emotional state, the desired emotional state and 20 suggestions of music from the user. m-Motion returned a playlist for each participant based on the songs suggested by the user. Adaptations of the Self Assessment Manikin (SAM) and Scherer's Semantic Emotional Space were used for the evaluation of the emotion achieved. The results suggest that when listening to the songs selected by m-Motion, the user approached the desired emotional state. There were 40 recommendations made by m-Motion, and in 95% of the recommendations, the users achieved the desired emotion in both evaluations.