SetMIR：将多兴趣检索视为集合预测问题

中文摘要 AI 辅助

基于嵌入的检索是工业推荐系统的核心，但单一用户嵌入往往过于局限，无法捕捉用户的多样兴趣。多兴趣检索通过使用多个用户嵌入解决这一问题，但现有方法仍存在两个问题：兴趣崩溃（不同嵌入学习到相同兴趣）和静态分配（即使部分嵌入不必要，服务阶段仍使用固定的检索预算）。我们提出SetMIR，将多兴趣检索视为集合预测问题。SetMIR使用Transformer编码用户的行为历史，并利用K个可学习查询解码一组用户兴趣，每个兴趣生成一个检索嵌入和一个存在分数。训练期间，匈牙利匹配为查询一对一分配目标，使匹配的查询学习不同兴趣，同时存在头学习哪些查询是活跃的。服务阶段，SetMIR利用存在分数和查询级别的非极大值抑制（NMS），仅发出活跃、非冗余的近似最近邻（ANN）查询。在Snap的动态产品广告（DPA）数据上，SetMIR在所有指标上均优于四种学习型多兴趣检索器，且每个请求发出的ANN查询减少33%。作为新的检索源部署在DPA生产栈中后，SetMIR使整体转化率（CVR）提升3.1%，相较于使用相同物品嵌入、ANN索引和检索配额的物品到物品检索源，点击率（CTR）提升44%、CVR提升51%。

英文摘要

Embedding-based retrieval is at the core of industrial recommender systems, but a single user embedding is often too limited to capture a user's diverse interests. Multi-interest retrieval addresses this by using multiple user embeddings, yet existing methods still suffer from two issues: interest collapse, where different embeddings learn the same interest, and static dispatch, where serving uses a fixed retrieval budget even when some embeddings are unnecessary. We propose SetMIR, which treats multi-interest retrieval as a set prediction problem. SetMIR encodes a user's behavior history with a transformer and uses K learnable queries to decode a set of user interests, each producing a retrieval embedding and a presence score. During training, Hungarian matching assigns targets to queries one-to-one, so matched queries learn distinct interests and the presence head learns which queries are active. At serving time, SetMIR uses presence scores and query-level Non-Maximum Suppression (NMS) to issue only active, non-redundant ANN queries. On Snap's Dynamic Product Ads (DPA) data, SetMIR outperforms four learned multi-interest retrievers on every metric while issuing 33% fewer ANN queries per request. Deployed as a new retrieval source in the DPA production stack, SetMIR lifts overall CVR by 3.1%, while lifting CTR by 44% and CVR by 51% over the item-to-item retrieval source with the same item embeddings, ANN index, and retrieval quota.

发表机构

• Snap Inc.（Snap公司）

机构由 AI 辅助整理，请以论文原文为准。