提升电商搜索语义召回，精准区分近似与互补商品。
Semantic Retrieval for Product Search in E-Commerce
- 双阶段训练：先用对比学习防近似商品误罚，再用偏好优化精调排序。
- 在多个业务品类和查询频次段均实现显著效果提升，线上实验验证显著。
- 适合需要高精度商品匹配与排序的电商平台研发团队参考。
电商语义检索需应对短、嘈杂且口语化的查询，在包含细粒度属性的大规模商品库中进行匹配。本文提出一种基于孪生大语言模型的双编码器，采用两阶段训练流程：第一阶段通过带假负样本掩码的对比学习，避免对近似商品的错误惩罚；第二阶段引入相对几率对齐（ROAR）偏好优化目标，将Bradley-Terry模型扩展至可变大小的分级相关性组，使用连续几率比边界进行优化。训练数据集也对应此流程设计：第一阶段使用替换查询-商品对提供粗粒度语义监督，第二阶段依赖分级相关性标注实现细粒度排序。所提系统不仅能准确召回精确匹配项，还能正确排序替代品与互补商品，在不同查询频率分层和业务垂直领域均取得提升，且经大规模在线A/B测试验证具有统计显著性。
原文摘要 · Abstract (English)
Semantic retrieval in e-commerce must handle short, noisy, and colloquial queries over large product catalogs with fine-grained attribute distinctions. We present a Siamese LLM dual-encoder trained through a two-stage pipeline: contrastive learning with a false-negative margin mask to prevent penalization of near-duplicate products, followed by Relative Odds Alignment for Retrieval (ROAR), a preference optimization objective that extends Bradley-Terry to variable-sized graded relevance groups via consecutive odds-ratio margins. The training corpus mirrors this progression - substitute query-product pairs provide coarse semantic supervision in Stage 1 and graded relevance annotations drive fine-grained ranking in Stage 2. The resulting system accurately retrieves exact matches while correctly ordering substitutes and complementary products, with gains confirmed across query-frequency strata and business verticals, and statistical significance validated through live A/B deployment at scale.
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。