Music Recommender Systems

• Tech Stack: Python, Spark, Latent Factor Model (LFM), Alternating Least Square (ALS) Implicit Feedback Model
• Github URL: Project Link

In this project, we implement a recommender system using data from ListenBrainz. We implement different recommender models to determine which one is able to produce better recommendations. We also analyze how long it takes to implement the LFM in different environments, i.e. Spark versus single-machine computation and see how spatial data structures can accelerate search at query time. The main focus of our analysis is the comparison between a baseline popularity model and the LFM. Ultimately, we find that the LFM produces better recommendations for users compared to the baseline popularity model.