Sentiment Analysis of Grab App Reviews with Machine Learning Approach

Abstract

Technological advances in online transportation services such as Grab facilitated user mobility. User reviews of the application were a valuable source of information for developers to improve service quality and for users to make decisions regarding service use. This research aimed to analyze the sentiment of Grab application user reviews using a machine learning approach. The system development method used in this research was the Agile method with the stages of Planning, Iterative Development, and Testing. The machine learning algorithms applied were Random Forest, Support Vector Machine (SVM), and Naive Bayes. The results of sentiment analysis of Grab application reviews were in the form of classification of reviews into positive, neutral, and negative sentiments. The test results showed that the Random Forest algorithm had the highest accuracy rate of 95.14%. This indicated that Random Forest was effective in identifying sentiment patterns in review data.