Classification Of Customer Satisfaction Levels For Tring Application Services At Pegadaian Syariah Using The C4.5 Algorithm

Abstract

Digital transformation in the financial sector encourages financial institutions to continuously improve the quality of digital services to meet customer needs and satisfaction. One of the digital services provided by PT Pegadaian Syariah is the Tring application, which offers various services such as pawn services, gold savings, financing, and online transactions. However, several issues are still encountered during its use, including feature maintenance and system instability, which may affect customer satisfaction. This study aims to develop a customer satisfaction classification model, identify the service factors that have the most significant influence, and measure the performance of the C4.5 algorithm in classifying customer satisfaction with the Tring application. A quantitative approach was employed using questionnaire data collected from 100 customers of the Plaza Taman Harapan Baru Branch of Pegadaian Syariah who had experience using the Tring application. The attributes considered consist of ease of use, service speed, feature completeness, application stability, information quality, and transaction quality. The classification process was performed using the C4.5 algorithm with the assistance of RapidMiner and evaluated using 10-Fold Cross Validation. The results show that Ease of Use (X1) was the most dominant factor, with the highest information gain value of 0.144965037, and was selected as the root node in the decision tree. The model achieved an accuracy of 83.00% and an Area Under the Curve (AUC) value of 0.931, which falls into the Excellent Classification category. These results indicate that the C4.5 algorithm provides good performance in classifying customer satisfaction into satisfied and dissatisfied categories.