JoMP: a mobile music player agent for joggers based on user interest and pace

Jogging is a common and cheap form of exercise which improves health and can reduce the risk of illness. Many people wear mobile devices to listen to music during this activity. However, the playlist in mobile devices are predefined and independent of the jogger's pace. Whenever the user wants to change the music that is playing, they have to stop exercising and push buttons on the device. In this paper, we introduce a novel service that provides the jogger with a smart music player agent to reduce the effort required to control the mobile device. This system involves music filtration and pace prediction technologies. An artificial neural networks that can filter out undesired music is used as the kernel of the music filter. Pace prediction is based on hidden Markov models (HMM), and similar pattern searching techniques are used to select music that has a suitable tempo. A series of experiments were carried out to demonstrate the performance of this system. The results show that this unique service is attractive to joggers.