Publication

Abstract : Online movie review platforms present a new and effective way for users to share feedback about movies and television shows. These platforms provide information to help the prospective audience make informed choices based on their personal preferences. By analyzing these reviews, production companies can also understand customers’ opinions on their projects. The analysis of sentiments in these reviews is used to generate an overall score. This score shows how the public perceives a movie. This research study identifies user sentiments by analyzing online user feedback on some movie review platforms. This research study considers the IMDB and Rotten Tomatoes dataset. The user sentiments are classified as either Positive or Negative by computing the Polarity score of the reviews. Finally, this research study examines the performance of different lexicon-based sentiment analysis tools like Textblob, VADER Sentiment Intensity Analyzer, SpaCy, and Textblob Naive Bayes Analyzer in detecting movie sentiments. Based on the accuracy of the predicted sentiment scores, this study prefers Textblob’s Naive Bayes Analyzer since it has delivered the best performance. The proposed model generates an accuracy of 73% and an F1-score of 0.78 for the IMDB movie reviews.