Stageboost：基于反事实估计的信号推荐
Stageboost: Recommending Signals Based on Counterfactual Estimation
浏览论文内容
中文总结 AI 辅助
该研究针对eBay商品详情页信号推荐问题，提出两阶段XGBoost模型，通过反事实估计优化信号配置，使整体GMB提升0.08%、配件GMB提升0.58%。
中文摘要 AI 辅助
信号是展示在eBay商品详情（VI）页面上的短文本或视觉片段，为用户提供所浏览商品的额外上下文信息，目的是促进智能购买并激励用户参与。本文提出一种基于两阶段XGBoost的模型，用于为商品详情页面合理配置信号。该方法在在线实验中，使整体商品成交总额（GMB）提升0.08%，零部件及配件成交总额提升0.58%，主要得益于高客单价商品转化率的提升。
英文摘要
Signals are short textual or visual snippets displayed on the eBay View-Item (VI) page, providing additional, contextual information for users about the viewed item. The aim of displaying these signals is to facilitate intelligent purchase and to incentivize engagement. In this paper, we present a 2 stage xgboost based model that optimally populates the VI page with signals. This approach has shown a 0.08% lift in overall GMB (Gross Merchandise Bought) and 0.58% increase in Parts and Accessories GMB, primarily due to increase in conversion of high average price items in online experimentation.
发表机构
- eBay（亿贝）
机构由 AI 辅助整理，请以论文原文为准。
