Improving Movie Review Sentiment Classification Through Advanced Machine Learning Models: A Comparative Study
Diaa Salama AbdElminaam, Amr Mohamed Mahmoud, Mohamed Ehab Anwar, Abdelrahman Abdelmageed Ahmed, Abdelrahman Mohamed Abdelfattah · 2024
Sentiment analysis plays a crucial role in understanding audience reactions to movies, offering valuable insights for filmmakers and industry professionals. This study explores the use of machine learning algorithms to classify movie reviews as positive or negative, aiming to improve the accuracy and efficiency of sentiment analysis. We compare the performance of four machine learning models: Logistic Regression, Support Vector Machines (SVM), Naive Bayes, and Random Forest, applied to the IMDB movie review dataset. Preprocessing steps, including tokenization and the application of Term Frequency-Inverse Document Frequency (TF-IDF), were used to convert text data into numerical form for model training. The models were evaluated based on accuracy, precision, recall, and F1-score. Our results show that SVM achieved the highest accuracy of 91.6%, outperforming other models. The findings highlight the potential of machine learning in enhancing sentiment analysis and provide a foundation for further research using more advanced techniques, such as deep learning, to improve classification performance.