Implementasi Algoritma Naive Bayes, Support Vector Machine, dan K-Nearest Neighbors untuk Analisa Sentimen Aplikasi Halodoc

Abstract

Received May 1, 2021 Revised May 25, 2021 Accepted May 28, 2021 During the Covid-19 pandemic, many people access information and even consult health problems online with the best doctors via smartphones. The Halodoc application is considered the most popular with 18 million users in 2020. So that many people have reviewed the application on the Google Play Store application provider. It may take a while to read the full review. However, if only a few comments are read, they are biased. For that, a platform is needed which can automatically identify positive or negative opinions. Sentiment analysis is a solution for the technique of classifying texts or sentiments into positive or negative opinion categories. The method used in this research is an experiment using the Naive Bayes algorithm, Support Vector Machine, and K-Nearest Neighbors. Evaluation is carried out using 10 Fold Cross-Validation. The results showed that K-Nearest Neighbors (KNN) had the best and most accurate performance in the sentiment classification because it produced the highest accuracy value of 95.00% and the largest AUC value of 0.985 compared to the Naive Bayes and Support Vector Machine algorithm.