arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2602.00899cs.LGcs.IR
收藏

面向领域适应和可扩展的密集检索用于基于内容的推荐

Domain-Adaptive and Scalable Dense Retrieval for Content-Based Recommendation

Aditya Birla Group(阿迪提亚·布尔拉集团)

Mritunjay Pandey

AI总结:

本文提出了一种基于双塔双编码器的可扩展密集检索系统,通过领域适应提升基于内容的推荐效果,实现高效服务和模型压缩。

摘要:

电子商务推荐和搜索通常依赖于稀疏关键词匹配(例如BM25),当用户意图与产品元数据存在词汇不匹配时,这种匹配会失效。我们将基于内容的推荐视为推荐作为检索:给定一个自然语言意图信号(查询或评论),通过语义相似性从大型目录中检索出前K个最相关项。我们提出了一种基于双塔双编码器的可扩展密集检索系统,该系统在Amazon Reviews 2023(时尚)子集上进行微调,使用监督对比学习和多负样本排序损失进行训练。我们从评论文本(作为查询代理)和物品元数据(作为正文档)构建训练对,并在50,000个采样交互上进行微调,最大序列长度为500个标记。为了提高服务效率,我们结合FAISS HNSW索引与ONNX运行时推理管道,使用INT8动态量化。在覆盖826,402个目录项的评论到标题基准测试中,我们的方法将Recall@10从0.26(BM25)提升到0.66,同时满足实际延迟和模型大小约束:6.1毫秒中位数CPU推理延迟(批量大小1)和模型大小减少4倍。总体而言,我们提供了一个端到端、可重复的蓝图,将领域适应的密集检索从离线训练转换为CPU高效的服务,适用于目录规模。

英文摘要:

E-commerce recommendation and search commonly rely on sparse keyword matching (e.g., BM25), which breaks down under vocabulary mismatch when user intent has limited lexical overlap with product metadata. We cast content-based recommendation as recommendation-as-retrieval: given a natural-language intent signal (a query or review), retrieve the top-K most relevant items from a large catalog via semantic similarity. We present a scalable dense retrieval system based on a two-tower bi-encoder, fine-tuned on the Amazon Reviews 2023 (Fashion) subset using supervised contrastive learning with Multiple Negatives Ranking Loss. We construct training pairs from review text (as a query proxy) and item metadata (as the positive document) and fine-tune on 50,000 sampled interactions with a maximum sequence length of 500 tokens. For efficient serving, we combine FAISS HNSW indexing with an ONNX Runtime inference pipeline using INT8 dynamic quantization. On a review-to-title benchmark over 826,402 catalog items, our approach improves Recall@10 from 0.26 (BM25) to 0.66, while meeting practical latency and model-size constraints: 6.1 ms median CPU inference latency (batch size 1) and a 4x reduction in model size. Overall, we provide an end-to-end, reproducible blueprint for taking domain-adapted dense retrieval from offline training to CPU-efficient serving at catalog scale.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑