This study aims to identify key factors necessary for improving product design by utilizing online review data and to present a Product Design Improvement Model that systematically structures these elements. To overcome the limitation of traditional product design research—often relying on surveys or expert evaluations that fail to fully reflect real consumer experiences—this study focuses on user-generated reviews as the primary data source.

Approximately 500 reviews of a specific fashion product were collected from the MUSINSA platform. Text mining techniques were then applied, including morphological analysis and stopword removal, followed by word frequency analysis to extract key keywords. Furthermore, a semantic network analysis based on word co-occurrence relationships was conducted to visualize consumers’ cognitive structures regarding the product.

The analysis revealed that consumers showed high interest in factors such as fit, material, design, price, and usability. By examining the interrelationships among these elements, major areas for product improvement were identified. Based on these findings, a data-driven design improvement model was developed, demonstrating its potential as a decision-making tool that reflects user experience (UX) in product planning and enhancement processes.

This study holds academic significance by quantitatively analyzing consumer experiences through unstructured text data and suggesting the applicability of data-driven approaches in design research. Additionally, it provides practical contributions by proposing a framework that enables companies and designers to incorporate consumer needs into product design through review data analysis.

Keywords

User Review, Text Mining, Natural Language Processing, Product Design and Design Improvement