User Profile Based Movie Recommender System Using Machine Learning

With the abundance of information online, users often struggle to make informed decisions. Recommendation systems (RMS) are crucial for helping users find relevant items among vast choices thereby using Content-Based Filtering and Collaborative Filtering algorithms. These systems are widely used across e-commerce, music apps, and streaming platforms like Netflix and YouTube to enhance user experience and drive business goals. Despite their benefits, RMS face some hazards such as cold start problems and scalability issues, which can impact the quality of recommendations. The proposed method uses the MovieLens-100k dataset, specifically the movies and ratings files, to recommend movies based on Content-Based Filtering. By preprocessing the data and calculating TF-IDF vectors and calculating the similarity for movie genres, the system builds user profiles and recommends movies with high similarity scores. Additionally, Single Value Decomposition (SVD) is used to address issues related to sparsity and scalability in collaborative filtering by reducing dimensionality.