PCap：Facebook Marketplace 个性化检索阶段多样性封顶
原标题：PCap: Personalized Retrieval-Stage Diversity Capping in Facebook Marketplace
AI 摘要
Meta 在 Facebook Marketplace 提出 PCap 个性化检索阶段多样性封顶框架，通过基于香农熵的用户多样性偏好建模、用户分桶和个性化类目上限，在多源候选召回阶段引入多样性约束。为调优高维参数，作者采用自动化在线优化方法 Parameter Tuning Sequence。大规模线上 A/B 实验显示 PCap 显著提升用户浏览体验相关的互动指标，并发现个性化多样性主要惠及偏好极端的用户。
正文节选
PCap: Personalized Retrieval-Stage Diversity Capping in Facebook Marketplace Abstract. We propose a personalized capping framework (PCap) to improve the diversity in Facebook Marketplace by introducing user-level diversity constraints at the retrieval stage. PCap models individual diversity preferences using Shannon entropy-based scoring, segments users into diversity buckets, and applies personalized category caps during multi-source candidate retrieval. To navigate the high-dimensional paramet