Researchers from Sultan Syarif Kasim State Islamic University in Riau, Muhammad Iqrom, along with four academics, analyzed the sentiments emerging from the app reviews of three of Indonesia's largest online transportation companies, namely Grab, Gojek, and Maxim in 2024.
In the journal "Analysis of Sentiments in Gojek, Grab, and Maxim Apps Using Support Vector Machine Algorithm", the researchers labeled three main sentiments: positive, negative, and neutral.
The results showed that neutral sentiments dominated all three apps, with Gojek having the highest percentage of 63.06%, followed by Grab at 43.83% and Maxim at 44%.
However, the largest positive sentiment was found in Maxim, at 42.45%, while Grab and Gojek recorded 32.83% and 20.21% respectively.
As for negative sentiments, Grab recorded the highest at 23.34%, followed by Gojek at 16.74% and Maxim at 13.55%.
"This illustrates the variations in users' perceptions of each app," the researchers wrote in the report cited on Tuesday (20/5/2025).
In addition to analyzing positive and negative sentiments, the researchers also revealed other sentiment variables. Overall, Maxim's strengths lie in its competitive pricing and driver professionalism.
The study also evaluated the performance of the sentiment analysis model. The results showed that Gojek recorded the best performance in sentiment classification with an accuracy of 94%, followed by Maxim at 90% and Grab at 87%.
The sentiment analysis of the Gojek, Grab, and Maxim apps was drawn from Google Play Store reviews. In the initial stage, review data from the three apps was collected using the Python programming language and the Google Play Scraper library.
The researchers used Python scraping techniques, which can be used to collect data over the past year, namely in 2024. As a result, the Gojek, Grab, and Maxim apps each received 1,800 rows of data.
The sentiment labeling was done manually, assisted by language experts, namely Dina Auliya, S.Pd, M.Pd Gr. The labeling was conducted on a total of 5,400 review data, consisting of 1,800 reviews each for the Gojek, Grab, and Maxim apps. The research results were published in March 2025.