<![CDATA[General Background: The rapid advancement of digital technology has catalyzed the proliferation of online service platforms, intensifying competition among providers to deliver optimal user experiences. Specific Background: In this landscape, food delivery services such as GoFood Merchant play a crucial role, yet user dissatisfaction remains a persistent challenge. Knowledge Gap: Despite the abundance of user-generated reviews, comprehensive sentiment and pattern analysis for GoFood Merchant remains limited, particularly in the integration of sentiment classification with pattern discovery and actionable recommendations. Aims: This study aims to analyze user sentiment toward the GoFood Merchant application using the Naïve Bayes algorithm and identify common complaint patterns via the Apriori algorithm, followed by solution formulation through the 5W+1H approach. Results: Utilizing 1,243 Play Store reviews, the sentiment classification model achieved an accuracy of 87%, indicating robust performance. Further analysis of negative reviews revealed five dominant keywords: “driver,” “order,” “aplikasi,” “resto,” and “iklan.” Novelty: The integration of sentiment analysis, association rule mining, and structured problem-solving provides a novel framework for understanding and addressing user dissatisfaction. Implications: The findings offer strategic insights for enhancing user experience and strengthening GoFood Merchant’s competitive advantage in the saturated online service marketplace. Highlights: Identifies dominant user complaints using data mining techniques. Combines sentiment classification with complaint pattern discovery. Provides actionable recommendations using the 5W+1H framework. Keywords: Sentiment Analysis, Naïve Bayes, Apriori Algorithm, User Complaints, GoFood Merchant]]>
