Analysis of algorithms and data processing efficiency in movie recommendation systems based on machine learning

Abstract

This study explores the use of content-based filtering and collaborative filtering algorithms in machine learning (ML)-based movie recommendation systems. The Collaborative Filtering and Content-Based Filtering algorithms work rather well, according to evaluation using Precision, Recall, and F1-Score metrics; Precision is approximately 0.82, Recall is approximately 0.85, and F1-Score is approximately 0.83. These findings show that both systems are capable of providing users with accurate and pertinent movie suggestions. The Collaborative Filtering and Content-Based Filtering algorithms work rather well, according to evaluation using Precision, Recall, and F1-Score metrics; Precision is approximately 0.82, Recall is approximately 0.85, and F1-Score is approximately 0.83. These findings show that both systems are capable of providing users with accurate and pertinent movie suggestions. The results demonstrate that both Collaborative Filtering and Content-Based Filtering produce highly accurate and relevant movie suggestions. When it comes to data processing, collaborative filtering is shown to be more effective than content-based filtering. The research advances the fields of information technology and computer science, especially in the creation of more precise and effective movie recommendation systems. The study also emphasizes how combining both algorithms in a hybrid approach could lead to even greater advancements in the creation of ML-based recommendation systems. Nevertheless, the study contains drawbacks, including the use of a small dataset and the failure to take into account additional variables that can affect movie choices. The goal of future studies should be to increase the dataset's size and take into account more aspects of the creation of movie recommendation systems.