Support Vector Machine Based Machine Learning for Sentiment Analysis of User Reviews of the Bibit Application on Google Play Store

Abstract

The increasing use of financial technology (fintech) applications has changed the investment patterns of users in Indonesia. Bibit, as one of the popular fintech investment platforms, receives many user reviews through the Google Play Store that reflect user perceptions and satisfaction levels. Although the volume of user reviews continues to increase, systematic analysis of user sentiment is still limited, making it difficult for developers to understand the needs and experiences of users. Therefore, an artificial intelligence-based approach is needed to efficiently and objectively extract and analyze user opinions. This study aims to conduct sentiment analysis of user reviews of the Bibit application using a Machine Vector Machine (SVM) based machine learning model. The research methodology includes data collection, pre-processing of texts, extraction of features using TF-IDF, as well as classification of sentiment into positive, negative, and neutral categories. Of the total review data, 7,801 data (79.99%) were used as training data, and 1,561 data (20.01%) were used as test data with a division ratio of 80:20 according to general standards in machine learning. The purpose of this study was to identify the dominant user sentiment and evaluate the classification performance of the SVM algorithm. The results of the experiment showed that the SVM model achieved high accuracy and was able to capture user opinions effectively, thus providing valuable input for developers in improving the quality of applications and user engagement on fintech platforms.