SENTIMENT ANALYSIS OF DIGITAL WALLET SERVICE USERS USING NAÏVE BAYES CLASSIFIER AND PARTICLE SWARM OPTIMIZATION

Digital wallet services provide many conveniences and benefits to its users. However, not all digital wallet service users have a positive opinion of the service. Sentiment analysis in this study aims to determine the opinions given by Dana and Isaku digital wallet service users whether they contain positive or negative opinions and apply the Naive Bayes Classifier and Particle Swarm Optimization (PSO) method to the sentiment analysis of digital wallet service users. The Naive Bayes Classifier method is used because it is simple, fast, high accuracy, and has good enough performance to classify data, but the Naive Bayes Classifier has the disadvantage that each independent variable is assumed to cause a decrease in the accuracy value. Therefore, this research added an attribute weighting method, namely Particle Swarm Optimization (PSO) to increase the accuracy of the classification of the Naive Bayes Classifier. This study uses data taken from Twitter as many as 490 tweet data. The test results using the confusion matrix and ROC curve show an increase in accuracy of the Naive Bayes Classifier Dana digital wallet from 60.00% to 91.67% and I.Saku digital wallet from 53.23% to 85.00%. T-Test and Anova test results show that the two classification methods tested have significant (significant) differences in Accuracy values.