Semantic Music Discovery Based On Rhythmic Similarity (RhyAn Recommends)

Abstract

With the main mode of music listening currently leaning towards streaming and digital media delivery in general, most services provide non-semantic music suggestion/discovery systems to expose users to new listening material. These recommendation systems mainly depend on data mining and metadata. With new music being released by artists every day, new music may lack the data or have ambiguous metadata thus delaying the recommendation to listeners. Whether the material is new or old, RhyAn Recommends goal is to give recommendations that are aligned with the individual listener’s taste. RhyAn recommends, RhyAn standing for “Rhythm Analyzer” counts semantic information extracted from audio signals, in this case rhythmic information. RhyAn is capable of extracting percussive sounds from music which RhyAn recommends uses to give listeners desirable recommendations