Sentiment Analysis Of Traveloka App Using Naïve Bayes Classifier Method

Abstract

Traveloka is currently the most popular startup in Indonesia with share traffic reaching 78.49% using smartphone and monthly visits whichreached 28.92 million based on a report in similarweb.com in May 2019. Traveloka, based on record, has been downloaded 10 million times since 2014with rating reaches 4.4 out of 5 stars. As of May 2019, there were 386,646 reviews from users in the PlayStore, ranging from positive and negativereviews. However, it is necessary to analyze with certain methods to summarize the review. Every review given will get a conclusion after collected, andsentiment analysis will provide user experiences from the Traveloka application within certain period. This research was conducted using the NaïveBayes Classifier method based on a review from the playstore to determine service quality. The purpose of this study is to find out the perceptions ofusers based on the measurement of service quality so that the results can be an evaluation for Traveloka in improving services. Studies show that duringthis period public opinion produced negative sentiments with Vmap value of 0.31020 greater than positive sentiment with a value of 0.16132.