用检索增强框架突破广告推荐长期记忆瓶颈，支持365天个性化历史数据。
Memento: Personalized RAG-Style Long-Retention Data Scaling for META Ads Recommendation
- 将用户行为当作文档库，用MMR检索相关历史交互，兼顾相似与多样。
- 在真实系统中实现10毫秒内响应，点击率和转化率分别提升0.25%和0.3%。
- 适合大规模个性化推荐场景，尤其需要长周期历史建模的工业级应用。
长期历史数据建模面临长上下文注意力稀释、系统效率低和灾难性遗忘问题，直接线性扩展如LastN方法无效。我们提出Memento，一种个性化检索增强框架，将历史用户行为视为文档语料库，广告请求作为查询，通过最大边际相关性（MMR）检索相关互动，平衡相似性与多样性。该框架包含两种互补应用：表示型Memento通过检索历史嵌入进行特征增强；数据型Memento则检索过往训练样本用于多轮训练。通过时序分块、INT8量化和异步服务的基础设施协同设计，Memento相比线性扩展实现5-10倍资源效率提升。系统每日处理请求延迟低于10毫秒，在点击率与转化率预测上分别获得0.25%-0.3%的归一化熵增益。上线后，于Facebook Feed和Reels带来1%的点击率提升和1.2%的转化率提升，成功将个性化建模扩展至365天以上的历史数据。
原文摘要 · Abstract (English)
Modeling of long history data suffers from long-context window attention dilution, system efficiency and catastrophic forgetting problems, where naive linear scaling approach like LastN would fail. We introduce Memento, a personalized retrieval-augmented framework that treats historical user engagements as a document corpus and ad requests as queries, retrieving relevant interactions via Maximal Marginal Relevance (MMR) to balance similarity with diversity. We identify two complementary applications: Representation Memento, which retrieves historical embeddings for feature augmentation, and Data Memento, which retrieves past training examples for multipass training. Through infrastructure co-design -- temporal chunking, INT8 quantization, and asynchronous serving -- Memento achieves 5-10$\times$ resource efficiency over linear scaling. Memento processes daily requests with sub-10ms latency, yielding 0.25-0.3% Normalized Entropy gain on both click-through and conversion prediction. In production, Memento delivers a 1% CTR lift on Facebook Feed and Reels and a 1.2% CVR lift, scaling personalization to 365+ days of history.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。