A Music Player with Song Selection Function for a Group of People

There are often situations in which a group of people gather and listen to the same songs. However, majority of existing studies related to music information retrieval (MIR) have focused on personalization for individual users, and there have been only a few studies related to MIR intended for a group of people. Here, we present an Android music player with a music selection function for people who are listening to the same songs in the same place. We assume that each user owns his/her favorite songs on his/her Android device. Once a group of users gathers each user can launch this player on his/her smartphone. Then, the player running on each device starts to communicate with other devices via Bluetooth. Information about songs stored in every device, along with the playback history, is collected to a device referred to as the master device. Then, the master device estimates each user’s preference for every song based on playback history and music similarity. The master device then extracts songs that are highly preferred and sends a command to start playback to the devices storing these songs. Our experimental results demonstrate the successful estimation of music preferences based on music similarity.