This study is purposed to evaluate the public service performance of the Immigration Office’s
M-Passport application using sentiment analysis. This study implemented the data of
comments or customer reviews from the Google Play Store. A web scraping technique was
employed to collect data, with a total number of 12,138 comments collected. Sentiment analysis
was applied to identify and label positive and negative comments. In order to ensure the
labeling results, a model test was carried out to confirm the accuracy level. The findings show
that 69.1 % comments are negative and 30.9% are positive. The model testing results indicated
a more than 80% accuracy level based on the Naive Bayes algorithm, Random Forest, Logistic
Regression, and Decision Tree. The results showed that the majority of users are not satisfied
with the service, indicating that the system performance of M-Passport has not satisfied user
expectations.