Music Playlist Recommendation Based on User Heartbeat and Music Preference

In this paper, we present a new user heartbeat and preference aware music recommendation system. The system can not only recommend a music playlist based on the user´s music preference but also the music playlist is generated based on the user´s heartbeat. If the user´s heartbeat is higher than the normal heartbeat which is 60-100 beats per minutes (age 18 and over) or 70-100 beats per minutes (age 6-18), the system generates a user preferred music playlist using Markov decision process to transfer the user´s heartbeat back to the normal range with the minimum time cost; if the user´s heartbeat is normal, the system generates a user preferred music playlist to keep the user´s heartbeat within the normal range; If the user´s heartbeat is lower than the normal heartbeat, the system generates a user preferred music playlist using Markov decision process to uplift the user´s heartbeat back to the normal range with the minimum time cost.