Sentiment Analysis of MobileJKN Application Reviews Using Neural Network Algorithm

Abstract

The advancement of information technology has encouraged the use of user data to improve digital services, particularly in health-related applications such as MobileJKN, developed by BPJS Kesehatan Indonesia. This research conducts sentiment analysis on user reviews of MobileJKN from the Google Play Store, aiming to identify key areas for improvement based on user perceptions. A Deep Learning approach is utilized, with Neural Networks as the primary model and Altair AI Studio as the main data processing tool. Following the Knowledge Discovery in Databases (KDD) methodology, the study involves various preprocessing stages including case folding, tokenization, filtering, stopword removal, and stemming, using the Kamus Besar Bahasa Indonesia (KBBI) to standardize local language terms. After preprocessing, clustering and classification are performed to extract sentiment patterns. The most frequently mentioned keywords “register,” “app,” “number,” “sign in,” and “verify” highlight common user concerns. The sentiment classification model achieved a 100% accuracy rate, with the Shuffled Sampling technique and a 90:10 training-testing ratio yielding optimal results. These findings demonstrate the effectiveness of Neural Networks in analyzing sentiment within health applications, providing valuable insights for developers seeking to enhance MobileJKN’s performance and user satisfaction. The study also offers a practical reference for future sentiment analysis research in the Indonesian digital health context.