Music recommendation service based on user impressions: Study of user interface acquiring appropriate impression words

Abstract

In recent years, with music streaming services becoming widespread, millions of songs have been made accessible on the Internet. However, music is not just entertainment but is used for various purposes in our lives, starting from the background music of shopping centers to music therapies. It is difficult to effectively select a single song from several songs based on a specific use scene. In this study, we propose a music recommendation method suitable for the elderly who have difficulty using advanced technology. We introduce a music database mapped to the emotional space, as well as interactive robots and agents, to achieve the appropriate atmosphere and users’ desired emotions. Finally, we consider a mechanism for obtaining appropriate impression (emotional) words from the elderly.