摘要
为提升社交媒体中服装类用户生成内容(UGC)的点击率,提出一种针对UGC封面的评价和测量方法。结合社交平台中UGC封面的特点,构建视觉特征和文本特征两个维度共13个关键特征的评价方案,应用图像处理技术适配并优化各特征的测量方法;以小红书平台4280个服装UGC封面为样本,通过负二项回归模型分析视觉特征对点击量的影响。结果表明,图片尺寸、模糊程度、前景位置、前景占比、人脸遮挡度、背景亮度、背景复杂度7个特征均显著影响消费者点击量。研究结果为社交媒体创作者的UGC创作活动及社交平台的UGC评价机制提供参考。
To improve the click-through rate of clothing user-generated content(UGC)on social media,this study proposed an evaluation and measurement method specifically for UGC covers.Incorporating the characteristics of UGC covers in social platforms,it constructed an evaluation scheme comprising two dimensions of visual features and textual features,with 13 key characteristics.Image processing techniques were applied to adapt and optimize the measurement methods for each feature.Taking 4280 fashion UGC covers from the rednote platform as samples,this paper analyzed the impact of visual features on click-through volume through a negative binomial regression model.The results indicate that seven features all significantly influenced consumer click-through rates,including image size,blur level,foreground position,foreground proportion,facial occlusion degree,background brightness,and background complexity.The findings provide practical references for social media creators'UGC production activities and the UGC evaluation mechanisms of social platforms.
作者
林琳
徐玉婷
何瑛
LIN Lin;XU Yuting;HE Ying(School of Fashion Design and Engineering,Zhejiang Sci-Tech University,Hangzhou 310018,China;Key Laboratory of Silk Culture Heritage and Products Design Digital Technology,Ministry of Culture and Tourism,Zhejiang Sci-Tech University,Hangzhou 310018,China;Apparel Engineering Technology Research Center of Zhejiang Province,Zhejiang Sci-Tech University,Hangzhou 310018,China)
出处
《服装学报》
北大核心
2025年第4期368-376,共9页
Journal of Clothing Research
基金
文化和旅游部重点实验室开放项目(2020WLB09)。
关键词
服装类用户生成内容
图像处理技术
视觉特征
封面图片
负二项回归
社会化电商
clothing user-generated content
image processing technology
visual features
cover picture
negative binomial regression
social e-commerce
