Collaborative Filtering based Hybrid Music Recommendation System

Even though people are nowadays listening all types of songs, still algorithms are struggling in many fields. With less historic figures, how does the system know the listeners like a new artist or a new song? And, how do you know which songs to suggest to new users? In this perspective, the proposed research work aims to assess the likelihood that a user will listen to the song repeatedly after the first apparent listening experience begins in the time frame. If the user has a recurring auditory occurrence within a month following the first apparent listening event, its goal is 1 and otherwise 0 identified. The approaches like collaborative filtering, content-based filtering, singular value decomposition (SVD) and factorization machines (FM) are also used. At last, the proposed system is hybridized by using SVD and FM.