Singing skills of each singer helps in choosing accurate songs for them. These skills are determined by using vocal
competence. Matching a song to singer’s vocal competence is the major and difficult task in song recommendation. There are
other song recommendation systems, which works on recommending songs of listener’s interest. These traditional approaches
did not consider the singer’s ability of singing which leads to poor singing performance by singer. This paper gives the
solution using song recommendation framework. This framework uses the term singer profile, which constitutes singer’s
vocal competence in terms of features of voice like pitch, intensity and also the voice quality of the singer. The paper then
gives technique to acquire singer profile and song profiles are stored in database. Then learning to- rank scheme is used to
rank the recommended song profile for each singer profile.
Keywords: Singing skills, Vocal Range Profile, Learning to-Rank.