SOCIOPHYSICS APPLICATION FOR SENTIMENT ANALYSIS OF CINEMA MOVIES USING NAÏVE- BAYES AND TEXTBLOB WITH PARTICLE SWARM OPTIMIZATION

• Main Author: Chusnul Amaliyah S., Almira
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/83769
• Institution: Institut Teknologi Bandung

Movie industry has developed rapidly and there could be big risks of investing in this industry. Predictions are needed to determine how long the movie should be shown and how many cinemas will show it. In this final project, sentiment classification and polarity analysis are carried out to determine the performance of a movie. In classifying the sentiment, two main methods are used: Naïve-Bayes method with Particle Swarm Optimization (PSO) and TextBlob. The results shown that the Naïve-Bayes method using Particle Swarm Optimization (PSO) could increase the accuracy of movie review sentiment classification in the first 4 weeks with an accuracy range of 80.70% to 84.99%. Predictions of movie performance cannot be seen directly only from the sentiment score data of movie reviews in the first 4 weeks. However, analysis using movie sentiment scores in the first 4 weeks can describe the popularity and public sentiment of movies quite well.