From User Preferences to Accurate Predictions: Enhancing Movie Recommendation Systems with Neural Collaborative Filtering and Sentiment Analysis

This paper introduces an advanced movie recommendation system that combines neural collaborative filtering (NCF) with sentiment analysis to improve personalization, accuracy, and robustness in recommending movies. By integrating user preferences with the sentiment polarity of reviews, the system refines recommendations, effectively addressing challenges such as the cold start problem, data sparsity, and limited semantic understanding. Unlike traditional collaborative and content-based filtering methods, which often struggle with diversity and accuracy, our approach utilizes deep learning within NCF to reveal hidden patterns in user behavior, enhancing recommendations’ relevance and precision. The proposed system is evaluated using standard performance metrics, including RMSE, MSE, and MAE, demonstrating its superior performance over conventional filtering techniques. Its enhanced scalability and adaptability position it as a promising tool for personalized content delivery in digital entertainment, with considerable potential for large-scale, dynamic recommendation environments. This research contributes to the existing knowledge on recommendation systems and offers new insights into improving content personalization and user satisfaction. The novelty and scientific value of this work lie in applying deep learning to tackle the challenges of accurate content recommendation in the rapidly evolving digital media landscape.