Research on cigarette packaging strategy based on Kansei engineering and decision tree

Le résumé fourni par la source

To accurately quantify consumer emotional needs and visual preferences in cigarette packaging design, and address the subjective and imprecise limitations of traditional design decision-making methods, this study integrates Kansei Engineering with Decision Tree. By utilizing web crawlers, 2360 consumer reviews from 2018 to 2024 were analyzed to extract representative emotional vocabulary. Additionally, 30 key design elements were summarized from 2359 packaging images. Through a Kansei-based questionnaire survey of 2891 users, 4061 valid rating records were obtained. CART decision tree models were then applied for layered analysis across different price groups and perceptual dimensions. The results revealed significant differences in consumer preferences across various styles and price ranges. For example, in the high-end style within the mid-to-high price segment, consumers significantly prefer modern illustrations and botanical background pattern (feature importance: 0.51 and 0.57, respectively). Meanwhile, in the luxury style within the low-to-mid price segment, geometric pattern textures and partial compositions are notably preferred (confidence level: 87.3%). This study innovatively establishes a data-driven design model combining Kansei engineering and decision-tree algorithms, providing a scientifically grounded theoretical reference for optimizing cigarette packaging design.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.