A BERTopic-Based Analysis of Consumer Responses to Men’s Cosmetics

• 발행기관: 한국식품보건융합학회
• 저자명: Seokjin IM
• 간행물 정보: 『식품보건융합연구』제11권 4호, 11~19쪽, 전체 9쪽
• 주제분류: 복합학 > 과학기술학
• 파일형태: PDF
• 발행일자: 2025.09.30

국문 초록

This study explores the rapidly growing men’s cosmetics market by analyzing male consumers’ review data on men’s cosmetics and usage experiences through text mining techniques. Unlike previous studies that relied on surveys, this research uses unstructured review data to better understand consumers’ needs expressed in natural language. A total of 5,193 reviews were collected from the Korean e-commerce platform Coupang. After preprocessing, the data was analyzed using keyword frequency analysis, Word2Vec-based association analysis, and BERTopic-based topic modeling. The most frequent keywords—such as “skin,” “price,” “all-in-one,” and “satisfaction”—suggest that men consider product functionality, skin compatibility, convenience, and emotional factors in their purchasing decisions. BERTopic modeling identified 30 topics, which were grouped into three main themes: functionality-centered, cost-effectiveness-centered, and emotional or social context-centered. The study also uncovered diverse consumption patterns, including preferences for all-in-one products, targeted solutions for specific skin concerns, and relational uses like gifting or shared family routines. These findings highlight how male consumers engage with cosmetics in multifaceted ways. By applying advanced text mining methods to real consumer data, the study contributes meaningful insights for both academia and industry, particularly in marketing strategy, product planning, and the design of personalized services for men’s cosmetics.

목차

1. Introduction

2. Literature Review

3. Research Methodology

4. Analysis Results

5. Conclusion

References

키워드

Cosmetics Consumer Response Text Mining BERTopic Topic Modeling