Sentiment Analysis of iPusnas Application Reviews on Google Play Using Support Vector Machine

Abstract

The iPusnas application is a digital library platform that allows users to borrow and read books online, free, and legally. Digital libraries provide literacy facilities without having to come to the library, take part in reducing the spread of the Covid-19 virus, and one of the government's efforts to increase reading interest in Indonesia. User experience in using the application usually conveyed in form of positive or negative reviews on Google Play which can be used to determine the quality of an application. However, too many reviews often make it difficult for readers to conclude the contents of the reviews. Based on that, this research aims to determine user sentiment of iPusnas application by conducting sentiment analysis using Support Vector Machine. The dataset used is iPusnas application reviews on Google Play with 6.946 clean data. Classification produces 94,24% accuracy, 92,38% precision, 83,86% recall, and 87,82% f1-score. This application has 75.1% positive sentiment and 24.9% negative sentiment with a list of frequently appearing words displayed in wordcloud. The final result of this analysis can be used as an evaluation to determine the steps for developing the application to be more optimal so that it can provide a better user experience.