Sentiment Analysis of Indonesia’s National Health Insurance Mobile Application using Naïve Bayes Algorithm

Abstract

Indonesia is developing one of the biggest single-payer social health insurance systems in the world, aiming to cover its entire population of over 260 million. To further expand membership and improve services, the National Health Insurance Administrator (BPJS Kesehatan) launched Mobile JKN, a mobile application that allows people to register, pay monthly contributions, set appointments with healthcare providers, and more. In this paper, sentiment orientation is explored considering the positive and negative sentiments using Indonesian user reviews for Mobile JKN application to discover user perception against service quality of Mobile JKN. Naive Bayes classifier method is used. The accuracy obtained from the model is 93%. Then the model is applied to analyze the sentiment of user review for Mobile JKN application version 3 and above. The results indicate that Mobile JKN version 3 and above tend to receive negative reviews rather than positive from users. Positive reviews do not refer to the specific features of the application. Whereas in negative reviews, many users complain about application updates that are too frequent and problems when logging in and signing in related to phone number verification.