ANALISIS SENTIMEN PADA ULASAN APLIKASI AMAZON SHOPPING DI GOOGLE PLAY STORE MENGGUNAKAN NAIVE BAYES CLASSIFIER

Abstract

Sentiment analysis or opinion mining is a study that analyzes people's opinions, thoughts and impressions on various topics, subjects, and products or services. The development of social media makes public opinion data available which can be found easily on the internet. The large volume of data causes the need for an automatic system to classify the data based on different aspects because classifying data manually is a time-consuming process. In this study, sentiment analysis will be carried out with a machine learning-based approach using the Naive Bayes algorithm using user review data on the Amazon Shopping application on the Google Play Store. The classification results using the four Naive Bayes algorithms produce an average accuracy of 82.15%, precision of 72.25%, recall of 83.49%, and f1-score of 77.41%. Multinomial NB produces the best accuracy among the four Naive Bayes algorithms used, which is 86.74%. The values of precision, recall, and f1-score are 78.82%, 85.90%, and 82.21%, respectively.