A Comparison of E-wallet Applications Opinion Based on Sentiment Analysis

Abstract

Nowadays, e-wallets have become a popular payment tool that can be used for various types of transactions. The rapid growth of e-wallets today is closely related to advancements in digital technology, like smartphones. Application X and Y are frequently used by people in Indonesia. With so many reviews about these e-wallets, we can notice that there are many positive and negative reviews. These reviews can be used for sentiment analysis of the applications and classified into positive or negative categories. 1,000 reviews collected from Google Play Store were divided into two sentiment classes: positive sentiment or negative sentiment. This research uses the Naïve Bayes classification method. The algorithm produces an accuracy of 91.87% for the testing model, 86% accuracy for sentiment analysis of Application X and 98% accuracy for sentiment analysis of Application Y. The results indicates that Application X is outperforms than the other one.