发表机构
Danggeun Market Inc. (Karrot)（당근마켓 주식회사（당근））
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本研究针对本地社区平台用户建模中对比学习的不可能负样本问题，提出区域约束批采样方法，可提升用户表示质量并优化推荐、广告排序，相关嵌入已部署生产。
AI 中文摘要
对比学习被广泛用于大规模推荐系统中的用户建模，标准的批内负样本隐含假设用户可接触到任意物品。但在Karrot这类本地社区平台中，接触受地理约束，许多用户-物品对因设计原因不可能被接触，却在训练中仍被当作负样本，稀释了对比学习信号。我们针对该不可能负样本问题，提出区域约束批采样（RCBS）这一简单有效的批处理方法，构建区域同构小批量，使用户主要与可行接触的物品进行对比。通过将不可能负样本替换为可行负样本，RCBS在现实接触约束下自然引入更难、更具信息性的负样本。经离线评估和在线A/B测试，RCBS持续提升用户表示质量，进而优化主页推荐排序、召回及展示广告排序，生成的用户嵌入已在各应用的生产环境部署。
英文摘要
Contrastive learning is widely used for user modeling in large-scale recommender systems, where standard in-batch negatives implicitly assume universal exposure that any user can be shown any item. On local community platforms such as Karrot, however, exposure is geographically constrained; many user-item pairs are impossible by design yet still treated as negatives during training, diluting the contrastive learning signal. We address this impossible negatives problem and propose Region-Constrained Batch Sampling (RCBS), a simple yet effective batching method that constructs region-homogeneous mini-batches so that users are contrasted primarily against items they could feasibly see. By replacing impossible negatives with feasible ones, RCBS naturally introduces harder and more informative negatives under realistic exposure constraints. With offline evaluations and online A/B tests, we show that RCBS consistently improves user representation quality and consequently enhances home feed ranking, retrieval, and display ads ranking. The resulting user embeddings have been deployed in production across various applications.
CommentsAccepted at SIGIR 2026 (Industry Track)
Journal refProceedings of the 49th International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR '26), pages 4639-4643, 2026
