Publication

Abstract : Movies significantly contribute to entertainment, inspiration, and an escape from reality. With the proliferation of online reviews, sentiment analysis facilitates a quick understanding of a film's reception. This study investigates sentiment classification using IMDb reviews, comparing Naïve Bayes models with Long Short-Term Memory (LSTM) networks. Specifically, Multinomial and Bernoulli Naïve Bayes classifiers were evaluated using Count Vectorizer and TF-IDF feature extraction techniques. The findings indicate that Bernoulli Naïve Bayes achieved an accuracy of 84.58%, while Multinomial Naïve Bayes attained 85.13%. However, the LSTM model outperformed both, reaching 88.17% accuracy. Beyond accuracy, LSTM also demonstrated superior precision