A naïve bayes sentiment analysis for fintech mobile application user review in Indonesia

Abstract

The growth of Fintech industries in Indonesia, as the 5th most internet users in the world, is tremendous, with a predicted of 16,3% year on year user growth and a total of $176,75 Million USD in investment on Fintech Startup it has become one of the biggest potential market in the world. Therefore, with this fast-growing market, Fintech Companies need to know their user opinion in Realtime in order to face their competitor on the market. Existing user review on Application Store already existed but there are no classification based on the subject of the review, because it only focus on the application not on the business level, therefore there are in need of a sentiment analysis to classify a user review based on the subject, in this paper we do a sentiment analysis using Naïve Bayes algorithm in Bahasa Indonesia and in English, while also seeing the significance of data cleansing on the data, the resulted accuracy of the test in this paper between the analysis is 78% for English and Bahasa, and 75% for Bahasa Only.