Enny Aryanny
Department of Industrial Engineering, Faculty of Engineering and Science, Universitas Pembangunan Nasional “Veteran” Jawa Timur, Surabaya, Indonesia

Published : 1 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 1 Documents
Search

Sentiment-Based User Satisfaction Analysis of Indonesian E-Grocery Applications Using Naïve Bayes and XGBoost Farista Lilmumazzaini; Enny Aryanny
AJARCDE (Asian Journal of Applied Research for Community Development and Empowerment) Vol. 10 No. 3 (2026)
Publisher : Asia Pacific Network for Sustainable Agriculture, Food and Energy (SAFE-Network)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.29165/ajarcde.v10i3.1284

Abstract

This study evaluated user sentiment and satisfaction across Alfagift, Klik Indomaret, and HappyFresh by analyzing 9,000 Google Play Store reviews collected between June 2024 and June 2026. The research applied text preprocessing, TF-IDF weighting, an 80:20 train-test split, and 5-fold cross-validation. Classification performance was statistically validated using McNemar’s test. Multinomial Naïve Bayes achieved 75% accuracy, 76% precision, 74% recall, and a 74% F1-score, while XGBoost reached 86% accuracy, 85% precision, 85% recall, and an 85% F1-score. Sentiment labeling identified 5,042 negative reviews (56%) and 3,958 positive reviews (44%), with Klik Indomaret recording the highest negative share at 64.4%, Alfagift at 56.4%, and HappyFresh recording the highest positive share at 52.8%. The proposed 5W+1H improvement priorities provide actionable insights for practitioners to repair application stability, synchronize real-time inventory, fix OTP authentication, and accelerate refund service level agreements (SLA), aligning with SDGs 8, 9, and 12. Contribution to Sustainable Development Goals (SDGs):SDG 9: Industry, Innovation, and InfrastructureSDG 12: Responsible Consumption and ProductionSDG 8: Decent Work and Economic Growth