A music recommendation system (MRS) serves as a solution to the overwhelming amount of information in the digital music realm. In this research paper, we tackle significant challenges faced by recommendation systems, namely, the Long Tail phenomenon, data sparsity, and the cold-start problem by employing a weighted hybrid approach. This hybrid approach combines collaborative filtering techniques based on both user preferences and item characteristics. Notably, our proposed system incorporates contextual information when generating music recommendations. We conducted experiments on a benchmark dataset and on synthetic data generated from a Music Portal application. The results we obtained demonstrate the system’s ability to accurately capture user interests by considering various factors, including a user’s historical preferences, their profile, item similarities, timestamps, and their social connections.