Technology in Indonesia has advanced rapidly, making many changes in all aspects of life, one of which is the online streaming aspect, namely the Disney+ Hotstar application. Now Disney+ Hotstar is available on tablets, smart TVs, computers, and smartphones accessed from various places and times. Disney+ Hotstar has thousands of hours of various Pixar, Marvel films, as well as exclusive Indonesian and various countries' series. Although Disney+ Hotstar has a variety of interesting films and features, it does not guarantee that users are satisfied using the application. Because users have different opinions and assessments, this point can be seen from user reviews available on the Google Playstore. The main purpose of this study was to determine the assessment or sentiment of user reviews of the Disney+ Hotstar application by analyzing it. The technique used uses the Naive Bayes algorithm. A total of 1000 review data were obtained on December 28, 2024 from the Google Playstore via Google Colab, then processed using RapidMiner. The dataset went through the cleaning and preprocessing stages to become 873 review data. There were 128 good reviews and 745 bad reviews. TF-IDF weighting was performed before classification using 873 datasets. The classification stage used a cross-validation system and applied the Naive Bayes approach. Testing from this study revealed the accuracy results of the Naive Bayes algorithm of 76.06%, precision of 34.12%, and recall of 67.97%.
Copyrights © 2025