Monitoring the Quality of PeduliLindungi Application based on Customer Reviews on Google Play Using Hybrid Naïve Bayes -Laney p' Attribute Control Chart

Abstract

Indonesia is battling the COVID-19 pandemic. One of the government's strategies to break the virus's transmission chain is to track digital contacts in Indonesia using the PeduliLindungi application. The Google Play comment section is where users can express their opinions about the app. User opinions discovered on Google Play can be used to perform sentiment analysis and quality evaluation. The Naïve Bayes classification can be used to identify how user opinions contain positive, neutral, or negative sentiments in user reviews of the PeduliLindungi app on Google Play. The p and Laney p' charts can be used for quality evaluation. Laney p' control chart is an attribute chart used to monitor the proportion of defects with large and varied sample sizes. The data used in this study is from April 1, 2020, to March 31, 2022. According to the sentiment analysis results of user reviews of the PeduliLindungi app on Google Play, there are more negative reviews than positive classes. The classification accuracy has an Area Under Curve (AUC) value of 89.05%. This result shows that the test data has good classification. The monitoring results using p and Laney p' charts based on ratings and user reviews of the PeduliLindungi app show that the processes are still not statistically controlled. These findings indicate that the app developer still needs to make improvements.