사용자 선호 기반 생성형 인공지능 학습을 적용한 흥미성 패키지 디자인 방법 연구 -퉁이(统一) ‘셴청둬(鲜橙多)’ 사례를 중심으로-

초록

As competition in the Chinese beverage market intensifies, “interest-driven” package design that captures consumer attention plays an increasingly important role in enhancing brand awareness and encouraging purchase behavior. However, most current beverage packages remain focused on functional information delivery, lacking emotional expression and visual distinctiveness. This study aims to enhance the visual appeal and market competitiveness of fruit juice beverages by proposing interest-driven package designs that reflect consumer preferences using Generative Artificial Intelligence (GAI) technology. The research analyzed consumer preferences regarding color, graphics, and structural elements of interest-driven design through literature review, grounded theory, consumer surveys, and expert FGI (Focus Group Interview) based on the top eight fruit juice brands in China in 2024. The strengths and weaknesses of Uni-President’s “Beibingyang Fresh Orange (鲜橙多)” packaging were evaluated using SWOT analysis. Subsequently, design proposals before and after AI learning were generated using GPT-4o, and expert evaluations along with consumer testing (n=150) were conducted based on five indicators: color, graphics, structural creativity, interest, and purchase intention. The results revealed that the post-learning AI-generated packages demonstrated significantly superior aesthetic quality, emotional expressiveness, and purchase intention compared to the original and pre-learning designs. The AI-based interest-driven package design process proposed in this study is expected to make a practical contribution to the development of emotion-centered packaging strategies focused on user experience in the beverage industry.

키워드