온라인 리뷰 기반 펫샴푸 제품의 고객 요구 분석: 토픽모델링, 감성분석 및 중요도-성과분석을 활용한 통합 분석
As the pet care market continues to expand, pet shampoo products have become an important pet care product category for regular skin and hygiene management, with growing consumer attention to product quality and safety. This study conducted an integrated analysis using Latent Dirichlet Allocation (LDA) topic modeling, sentiment analysis, and Importance-Performance Analysis (IPA) to derive customer requirement attributes for pet shampoo products and identify product improvement priorities based on online review data. A total of 14,072 pet shampoo reviews were collected from Naver Shopping. After preprocessing, 7,103 reviews were used for sentiment analysis and 5,405 reviews were used for LDA topic modeling. Topic proportions derived from LDA were used as a proxy for importance (review salience), and normalized sentiment scores by topic were applied as the performance measure, which were then integrated into an IPA matrix. The analysis identified five customer requirement attributes: user experience and usability, daily preventive skin care, ingredient safety and reliability, skin condition care, and fragrance and coat condition. Daily preventive skin care and skin condition care were positioned in the high-salience, low-performance quadrant, suggesting potential priority areas for product improvement. This study applies a data-driven integrated analytical approach to customer requirement analysis based on online reviews and provides practical implications for supporting consumer-centered new product development and research and development (R&D) decision-making.