Comparing Machine Learning Models for YouTube Movie Trailer Comments: An Approach for Accuracy and Overall Sentiment Prediction

Abstract

This study compares multiple Machine Learning (ML) models for analyzing the sentiment of YouTube comments on movie trailers. The aim of this study is to determine which Machine Learning (ML) model can best accurately predict the overall sentiment of YouTube comments. We compiled a dataset of YouTube comments on a well-known movie trailer and labeled them based on their sentiment using a tokenizer. We then evaluated the performance of different ML models such as Naive Bayes, Support Vector Machine, k-Nearest Neighbors, Random Forest, and Bagging. Our findings show that the Naive Bayes model achieved the highest accuracy for sentiment analysis and provided the most accurate prediction for the overall sentiment of the comments.