Performance Evaluation of Movie-based Recommendation Systems using Hybrid Machine Learning Models

The recommendation systems are used in different sectors to offer tailored services for users. The major applications of these systems are in the entertainment sector, where the movie-based recommendation systems can suggest movie recommendations personalized to users based on their feedback or search history. The research study mainly focusses on evaluating the performance of recommendation models including content filtering and collaborative filtering methods like KNN model, SVD, Boost and the hybrid models that combine these approaches. Also, the study assesses the Matrix Factorization Technique as it ensures the robustness. The study compares the accuracy, efficiency, and robustness of each model using Netflix dataset and evaluation metrics designed specifically for movie-based recommendation systems, The result outcome shows that the hybrid models outperform in terms of accuracy and efficiency, and that the Matrix Factorization Technique is particularly robust. The results help academics and practitioners for their use case by illuminating the advantages and disadvantages of each model in the context of movie-based recommendation systems.