Music Recommendation System based on Big Data using Pyspark

In the evolving landscape of digital music consumption, personalized song recommendation systems play a pivotal role in enhancing user engagement and satisfaction. This article presents an in-depth exploration and analysis of an innovative song recommendation framework designed to address the challenges of delivering tailored music experiences. Utilizing advanced machine learning techniques, especially K-Means Clustering Algorithm and data analytics, the proposed system aims to offer song suggestions according to user preferences. The proposed system introduces a hybrid song recommendation system that combines the two distinct approaches i.e. Content and Collaborative based filtering. Various algorithms were implemented for Content based filtering whereas Hadoop and Pyspark was used for implementation of Collaborative filtering. Wide research and Evaluations based on various metrics were considered. The research also draws the strengths and drawbacks or limitations of Content based Recommendation over Collaborative based Recommendation System. The results highlight the system's capability to overcome limitations eventually enhancing user experience by offering a better experience.