Movie Recommendation Exact Prediction using Machine Learning Approaches
E. Vengatesh, P. Shalini, Misma Fredic, K. Ramachandra Raju, G. Uma Maheswari, R. Saravanakumar · 2025
As of right now, the recommendation system has made it easier to find the things you need. Movie recommendation systems help movie buffs by suggesting movies to watch, removing the time-consuming and confusing process of choosing among thousands or even millions of possibilities. As a result, a recommendation system is necessary to handle this enormous amount of data and quickly extract useful information that is in line with the user’s preferences. The approach for a movie recommendation system that uses Cosine Similarity in Large Language Models to propose similar films based on the user’s chosen film is described in this article. Even if today’s recommendation algorithms work well, they are unable to determine whether a film is worth the time spent. This approach improves the user experience by using machine learning to do sentiment analysis on reviews of the chosen film. To improve accuracy and efficiency, two supervised machine learning techniques are employed: the Random Forest (RF) Classifier and the XGBoost Classifier. This research compares RF with XGB using criteria including F1 Score, Accuracy, Precision, and Recall. RF had an accuracy score of 80%, while XGB had a score of 81%. As such, XGB outperforms RF and has better appropriateness for sentiment analysis.