Perbandingan Analisis Sentimen Ulasan Aplikasi Ajaib Kripto Menggunakan Metode Naïve Bayes dan K-Nearest Neighbor

Abstract

Developments that occur in the investment sector make people interested in investing. The platform that is often used is the crypto magic application on the Google Play Store. Potential users see app reviews, which are based on user opinions and used as consideration before deciding to use the app. So data analysis is needed. One way of analyzing data is called sentiment analysis. There are many methods that can be used for sentiment analysis. So the author uses the Naive Bayes and K-Nearest Neighbor methods to find out a more accurate method. The data taken for this study is a review of the crypto magic application collected from 2022 to 2024 with 4784 data. Dataset through the stages of data division using k-fold cross validation with k=5, then the preprocessing stage, TF-IDF transformation, training Naive Bayes and KNN models with training data. The model was then tested on test data. The result is that the Naïve Bayes method provides an accuracy rate of 82%, precision of 83%, recall of 98% and an f1-score of 90%. While the KNearest Neighbor method provides an accuracy rate of 80%, precision of 84%, recall of 94% and f1-score of 89%. The conclusion of this study is that the Naïve Bayes method has a better level of accuracy than the K-Nearest Neighbor method.