Comparison of The Results of The Naïve Bayes Method and Synthetic Minority Over Sampling Technique in Sentiment Analysis of User Reviews

Abstract

The rapid development of information technology encourages a variety of applications available on the Google Play Store. with various application categories, such as business, communication, education etc. One example of such an application is an online course application, namely Skill Academy from Ruangguru, which offers a variety of online guidance in the fields of education, self-development, and career. So from this, sentiment analysis will be carried out to understand a person's opinion and attitude towards a particular subject, theme, or entity in a text on the Skill Academy application from Ruangguru. This research aims to compare the performance of the Naı¨ve Bayes classification algorithm with the Syntetic Minority Over-Sampling Technique (SMOTE) on the sentiment of the Skill Academy application. This study shows the results of calculations without SMOTE and compares them with the results of SMOTE calculations. The results of this study are also expected to provide a better understanding of analyzing dataset imbalance on sentiment analysis results using Naıve Bayes and SMOTE techniques. In addition, it can be used in conjunction with appropriate evaluation methods to produce a more accurate model.