Measuring contact tracing service quality using sentiment analysis: a case study of PeduliLindungi Indonesia

This research aims to determine the service quality dimensions of contact tracing applications based on user perceptions which are still lacking in the literature. This research employs a text mining approach, specifically topic modeling and sentiment analysis, on the user reviews of the PeduliLindungi contact tracing application. The process of data collection, pre-processing of the reviews, grouping of the reviews into topics, measurement of the score for each service quality dimension, and evaluation of the total score of the PeduliLindungi service quality was carried out. This research revealed four main dimensions of service quality for the contact tracing application: system efficiency, functional benefit, system availability, and emotional benefit. The service quality score was 66.5% for system efficiency, 54.4% for functional benefit, 51.5% for system availability, and 46.2% for emotional benefit. Based on the user perspective, the research highlighted that system efficiency and functional benefit were the most crucial factors in determining the service quality of a contact tracing application. The government needs to pay more attention to the emotional benefit dimension, which received the lowest score of 46.2%. Users reported feeling less excitement and encouragement from the service provided. To improve the service quality of PeduliLindungi, the government can address the users' problems, such as app crashes, unresponsive displays, inconsistent availability of vaccine certificates, and inadequate app functionalities.