This study tries to analyze the sentiments that exist in the Bibit application review by applying the Naïve Bayes, Support Vector Machine (SVM), C4.5, K-Nearest Neighbor (KNN) methods. The results of the accuracy, recall, precision values will be a refere

Antika Zahrotul Kamalia (Universitas Pelita Bangsa)
Andi Al Zaroni (Universitas Pelita Bangsa)
Miftah Wangsadanureja (Universitas Pelita Bangsa)

Jurnal Ilmiah SIGMA: Informatics Engineering Journal of UPB
Vol 13 No 1 (2022): Maret 2022

This study tries to analyze the sentiments that exist in the Bibit application review by applying the Naïve Bayes, Support Vector Machine (SVM), C4.5, K-Nearest Neighbor (KNN) methods. The results of the accuracy, recall, precision values will be a reference in determining which method is the most appropriate in analyzing the sentiment of Bibit applications. In this study, we tested 464 datasets which were divided into 310 positive sentiment data and 154 negative sentiment data. The evaluation of the model uses 4 Fold Cross Validation, and the first test does not use Particle Swarm Optimization and the second uses Particle Swarm Optimization. In this study, the Naïve Bayes algorithm has the highest accuracy, recall and precision values in analyzing the sentiment analysis of the Bibit application. Keywords : Sentiment analysis, Naïve Bayes, Support Vector Machine(SVM), C4.5, K-Nearest Neighbor(KNN), Particle Swarm Optimization, Bibit.