跨国语码混合的生成式推荐框架 CMRec
原标题:Cross-Country Code-Mixing for Generative Recommendation
AI 摘要
该论文提出 CMRec,一个面向跨国电商场景的生成式推荐框架。由于各国用户与商品 ID 空间彼此独立,现有生成式推荐仅在参数层面共享知识,数据层面缺乏跨国监督。CMRec 借鉴多语言 NLP 的语码混合思路,先学习跨模态内容与行为共现的共享语义码本,再通过满足静态(内容)与动态(价格、受众、热度)双重约束的 token 级替换合成跨国序列,并引入上下文感知损失对混合样本重新加权。在两个真实跨国数据集和一次在线 A/B 测试中,CMRec 提升了数据稀疏国家的推荐质量,同时保持数据丰富国家的表现,在大规模电商平台上实现广告收入 +1.77%、订单量 +2.64%。
正文节选
Cross-Country Code-Mixing for Generative Recommendation Abstract. Cross-country recommendation on modern e-commerce platforms is typically deployed with disjoint user and item ID spaces across markets, removing the shared anchors that conventional cross-domain methods rely on. Generative recommendation (GR) mitigates this by mapping items into a shared token space and training a unified model, but existing approaches keep behavior sequences strictly country-specific, so knowledge transfer occurs