Human-Computer Interaction in E-Commerce: A Study of Tokopedia’s Usability, Efficiency, and Satisfaction for Elderly Consumers

Alvianus Alvianus Dengen (Electrical Engineering Department, Universitas Teknologi Sulawesi, Makassar)
Rahman Radjab Do Esa (Electrical Engineering Department, Universitas Teknologi Sulawesi, Makassar)
Budiawan (Electrical Engineering Department, Universitas Teknologi Sulawesi, Makassar)

Journal of Computer and Information Systems Ampera
Vol. 6 No. 3 (2025): Journal of Computer and Information Systems Ampera

Introduction: Elderly individuals are increasingly engaging with e-commerce platforms like Tokopedia. However, cognitive and technological barriers may affect their user experience. This study aims to analyze elderly user perceptions of Tokopedia using sentiment analysis of user reviews. Method: The data were collected from elderly users’ reviews on Tokopedia. Sentiment classification into positive, negative, and neutral categories was conducted using machine learning algorithms. Model performance was evaluated using precision, recall, and f1-score metrics, presented through heatmaps and pie charts. Results: The analysis revealed that 58.0% of reviews were negative, 34.8% positive, and 7.2% neutral. The classification model yielded a high f1-score for the negative class (0.91) and perfect precision for the positive class (1.00), but relatively low recall for the positive sentiment (0.71), suggesting challenges in identifying subtle satisfaction expressions from elderly users. Discussion: These findings indicate persistent barriers in the online shopping experience for elderly users, particularly in terms of interface and usability. Recommendations are proposed to improve age-friendly design and platform accessibility.