<![CDATA[This study analyzes customer satisfaction with customer service using data mining techniques and the Decision Tree algorithm. The data was obtained from customer questionnaires completed after transactions and were processed through pre-processing, attribute labeling, and missing value handling. The dataset was split into 80% training data and 20% testing data to build and evaluate a classification model with two target categories: satisfied and dissatisfied. The modeling results show that the Consideration Label is the most dominant factor in determining customer satisfaction, while the Suggestion Label serves as a supporting attribute. Model evaluation produced an accuracy of 58%, with precision, recall, and F1-score for the dissatisfied class of 0.62, 0.66, and 0.64, respectively, and for the satisfied class of 0.53, 0.49, and 0.51, respectively. Based on these results, the Decision Tree method can be used to classify customer satisfaction, although further improvement in model performance is still needed to obtain more optimal predictions.]]>
Copyrights © 2026
