ANALISIS SENTIMEN APLIKASI BRIMO PADA ULASAN PENGGUNA DI GOOGLE PLAY MENGGUNAKAN ALGORITMA NAIVE BAYES

Abstract

The development of technology is currently growing rapidly including applications in the field of banking which is currently widely used for mobile transactions without the need to bank. Of course, it is very easy for customer users to transact using the bangking mobile application.Reviews are an important source of information for developers to find out complaints from users or customers. user reviews that contain comments or reviews and user ratings are needed to improve the development of application performance. However, not all user reviews that represent opinions on the application of this research aims to analyze the sentiment of users who use the brimo application on google play with data retrieval through web scraping techniques. Sentiment analysis is the process of evaluating whether a text has positive or negative sentiment. Brimo application users on google play is one of the platforms widely used by BRI customers for banking transactions. The method used in sentiment analysis uses Naive Bayes algorithm. from the results of research on the sentiment analysis of brimo application users into positive and negative categories where data were taken through scraping results from August 2022 to January 2023 as many as 1550 text data, it was found that application users gave positive reviews where from the number of predicted positive sentiment classification results as many as 1012 and negative sentiment as many as 894 text data means that users leave many positive comments on the brimo application, but not a few users also leave negative comments on the brimo application. And classification results using naive bayes method obtained acuracy of 84.52% precision 82.51% and recall 87.62%.