【摘要】 针对现有服装个性化推荐模型存在精度低、用户满意度不高以及个性化指标不完善等问题，通过构建包含用户主客观特征的完整个性化指标体系，利用时尚专家服装穿搭意见与用户历史行为数据等，建立基于用户客观特征的层次分析法(Analytic Hierarchy Process,AHP)推荐模型和基于用户主观特征的K-Means推荐模型，并通过融合2个模型以平衡客观和主观因素的影响，进一步提高多特征融合推荐模型的综合性能。经过对比验证，该融合模型精度与用户满意度表现优异，且具有较强的鲁棒性和可解释性。更多还原

【Abstract】 Addressing issues such as low accuracy,insufficient user satisfaction,and imperfect personalization indicators in ex-isting clothing personalization recommendation models,a comprehensive personalization indicator system incorporating both subjec-tive and objective user characteristics was proposed. By leveraging fashion experts’ advice on clothing coordination and users’ histor-ical behavior data,an AHP(Analytic Hierarchy Process) recommendation model based on users’ objective characteristics and a K-Means recommendation model based on users’ subjective characteristics were developed. By integrating these models,the influence of both objective and subjective factors was balanced,further enhancing the overall performance of the multi-feature recommendation model. After comparative validation,the model demonstrates superior accuracy and user satisfaction compared to other models,along with strong robustness and interpretability.更多还原