<![CDATA[In realizing the Electronic-based Government System (Sistem Pemerintahan Berbasis Elektronik - SPBE) policy, the National Civil Service Agency (Badan Kepegawaian Nasional - BKN), as the fostering government agency for the State Civil Aparatus (Aparatur Sipil Negara - ASN), needs to carry out an accurate, up-to-date, and integrated data management through an Android-based application named MySAPK. Over time, users encounter various troubles operating this application and writing down their reviews on the Rating & Review feature in Google Play Store. From May 9, 2017, until October 18, 2021, reviews written by users amounted to 4,778. This paper conducted a sentiment analysis on MySAPK users’ reviews. Stages carried out included data collection, data labeling (annotation), data preprocessing, word feature extraction, classification modeling, modeling evaluation, sentiment analysis, and recommendation result preparation. The classification modeling of the sentiment using the naïve Bayes and support vector machine (SVM) resulted in an accuracy level of 92.47% and 94.14%, respectively. The sentiment measurement results showed that users who wrote reviews with positive sentiments amounted to 2,118 (44.3%), and negative sentiments amounted to 2,660 (55.7%). Factors that prompted users to leave positive sentiment reviews are the excellent application quality, providing benefits, making it easier to fill out and store ASN data, and gratitude comments for BKN. On the other hand, factors causing users to leave negative sentiments reviews include requesting to fix the application, having difficulty accessing the application, failing to fill out and update data, and encountering an error server. To address these issues, this paper suggests that BKN could increase supporting server capacity and update the most recent version to fix the bugs. The research result is expected to serve as a reference for BKN in evaluating and improving the quality of ASN services via the MySAPK app.]]>
Copyrights © 2022
