아마존 이커머스 리뷰 빅데이터를 활용한 화장품 만족·불만족 요인 탐색 -TF-IDF 및 LDA 분석을 중심으로-

Exploring Factors of Satisfaction and Dissatisfaction in Cosmetics Using Amazon E-commerce Review Big Data
  • 하오선
  • 신혜원

초록

This study aims to explore the factors of satisfaction and dissatisfaction among overseas consumers by analyzing item-specific best-selling cosmetics on Amazon, as the importance of online cosmetics markets grows alongside the global e-commerce market. Using Selenium and BeautifulSoup, review data on Amazon’s bestselling cosmetics were collected, and TF-IDF and LDA topic modeling were applied to identify the factors consumers consider significant when evaluating cosmetics. According to the TF-IDF analysis, consumers were highly satisfied with the moisturizing effects, softness, and wrinkle reduction of night creams, but some with sensitive skin reported side effects and expressed dissatisfaction. For essences, satisfied consumers highlighted the product’s refreshing and lightweight texture, deeming it suitable for morning skincare routines. In contrast, dissatisfaction often stemmed from complaints about packaging and price rather than side effects. Lipstick reviews revealed that consumers were pleased with its longevity, value for money, color, and overall feel. However, negative feedback centered on packaging seal issues, stickiness, and perceived poor value. For foundation, satisfied customers appreciated the product’s smooth application, coverage, natural finish, and color options. Many praised its ability to provide ample coverage while maintaining a lightweight feel. However, dissatisfaction arose due to dryness, heaviness, oiliness, or mismatched colors, with some also expressing disappointment over the packaging and price-performance ratio. Through LDA topic modeling, consumer reviews on cosmetics revealed six main topics: 'skin irritation reactions,' 'cream moisturizing effects,' 'lip product color and longevity,' 'scent and ingredients of liquid products,' 'tone correction and coverage,' and 'packaging quality and management.' These findings can be utilized as foundational data for developing tailored marketing strategies within the global online cosmetics market.

키워드

Big DataCosmeticsReviewSatisfaction・DissatisfactionText Mining빅데이터화장품리뷰만족・불만족테스트마이닝
제목
아마존 이커머스 리뷰 빅데이터를 활용한 화장품 만족·불만족 요인 탐색 -TF-IDF 및 LDA 분석을 중심으로-
제목 (타언어)
Exploring Factors of Satisfaction and Dissatisfaction in Cosmetics Using Amazon E-commerce Review Big Data
저자
하오선신혜원
DOI
10.23351/kscc.14.3.202412.004
발행일
2024-12
저널명
한국화장품미용학회지
14
3
페이지
471 ~ 482