PERBANDINGAN KINERJA METODE NAÏVE BAYES CLASSIFIER DAN K-NEAREST NEIGHBOR PADA ANALISIS SENTIMEN ULASAN MOBILE BANKING JENIUS

One of the impacts of the industrial revolution 4.0 is competition between banks and fintech, so that banks are not left behind in innovating by creating mobile banking, so that in Indonesia the number of mobile banking users is increasing every year and there are more and more customers so that banks need to pay attention to customer satisfaction. Sentiment analysis is one of the solutions that can be used to see customer satisfaction from the opinions of mobile banking application users. The test of 2000 data is divided into 1600 training data and 400 test data. The results of the analysis show that the K-Nearest Neighbor method is superior in analyzing sentiment and the results of the performance confusion matrix show superior K-Nearest Neighbor accuracy with an accuracy of 84 .06% and 83.06% while the Naïve Bayes Classifier obtained accuracy results of 83.06% and 82.56%. Keywords: mobile banking, Naïve Bayes Classifier, K-Nearest Neighbor, performance, sentiment analysis