| abstract |
| Background Recommender systems provide personalized experiences across various platforms. However, even in advanced algorithmic environments, issues such as user fatigue, diminished trust, and lack of persuasive recommendation logic persist. These challenges suggest that contextual alignment with user expectations is as important as technical precision. This study empirically examines the interaction between recommendation types and platform consumption contexts to identify user acceptance patterns.
Methods Three recommendation types―content-based, collaborative, and context-aware―were used as independent variables. Platform types were categorized according to consumption orientation (hedonic, exploratory, functional, symbolic) and used as moderating variables. Four user experience factors―satisfaction, perceived usefulness, fatigue, and continuance intention― were measured. A repeated measures ANOVA with Bonferroni post-hoc analysis was conducted with 200 participants, and in-depth interviews were performed to supplement the quantitative data.
Result Content-based recommendations received the most favorable evaluations across all platforms. Collaborative recommendations showed low acceptance due to unclear rationale and increased fatigue. Context-aware recommendations were effective on functional and symbolic platforms, but less so on emotion-driven platforms.
Conclusion The findings suggest that recommendation acceptance is influenced more by interpretability and trust than by performance alone. This highlights the importance of contextual alignment in personalized UX design. By analyzing cross-effects between recommendation approaches and platform types, this study provides a practical framework for future ethical design and AI-based UX strategies. |
|
|
| Key Words |
| 개인화, 추천 시스템, 플랫폼 유형, Personalization, Recommendation System, Platform Categories |
|
|
|
 |
|
Full Text