The Influence of Anthropomorphism and Presence on the Acceptance Intention of Makeup Image Optimization Recommendation Systems Utilizing Deep Learning Technology: Focusing on the Mediating Effect of Cognitive Absorption

This study was conducted to understand the relationship between HCI (Human-Computer Interaction) factors, user cognitive absorption, and acceptance intention in makeup image optimization recommendation systems. A preliminary survey and main survey were carried out targeting adults in their twenties, and various statistical methods including frequency analysis, exploratory and confirmatory factor analysis, and structural equation modeling were used to analyze the data using SPSS and AMOS programs. The results of the study confirmed that the survey items consistently and reliably measured each variable through reliability and validity analysis, and the model’s high fit was proven through principal component analysis and confirmatory factor analysis. Correlation analysis showed significant positive relationships between all variables, and most paths in hypothesis testing through structural equation modeling were statistically significant. In particular, a significant partial mediating effect was found in the path where anthropomorphism affects acceptance intention mediated by curiosity. This means that HCI factors such as anthropomorphism and presence positively influence the acceptance intention of the recommendation system through various aspects of cognitive absorption, such as user engagement, temporal dissociation, curiosity, and pleasure. These results provide important implications for user-centered service design and improvement, suggesting that it is possible to move towards increasing user satisfaction and loyalty by accurately reflecting individual user needs and preferences. Additionally, the findings promote a deeper understanding of technological innovation in the beauty industry and user experience design, laying the foundation for effective user-centered service development.