摘要
Abstract
The text layout of e-commerce posters is crucial for influencing consumer decision-making.However,traditional eye-tracking studies,constrained by small sample sizes(n≤50),struggle to reveal universal patterns.To address this,our research employs machine learning methods to conduct a quantitative analysis based on a sample of 60,548 posters,thereby overcoming the limitations of traditional sample sizes.This study is the first to systematically reveal the"T-shaped"concentrated distribution and"dual-intensive area"structure of text layouts.It further provides an in-depth analysis of the significant layout differentiation across categories(e.g.,food,home appliances,and 3C digital products)caused by differences in product forms and cognitive needs.The findings not only provide large-sample empirical support for Gestalt psychology and the F-shaped browsing pattern but also derive a set of highly practical,category-specific design strategies with precise parameters(such as font size,RGB color values,and position coordinates).This provides a data-driven scientific basis for e-commerce poster design,aiming to directly optimize information efficiency and conversion rates.关键词
电商海报/文本布局/机器学习/品类差异/视觉传达Key words
E-commerce poster/Text layout/Machine learning/Category difference/Visual communication分类
社会科学
