QueryAgent-R1: Bridging Query Generation and Product Retrieval for E-Commerce Query Recommendation

2026-06-03

AI Summary

This work addresses a critical limitation in existing e-commerce query recommendation methods, which prioritize query relevance at the expense of alignment with users’ downstream conversion preferences, often yielding high click-through but low conversion rates. To bridge this gap, we propose QueryAgent-R1, a memory-augmented agent framework that integrates query generation and product inventory retrieval in a closed-loop manner. Our approach introduces a consistency-based reward mechanism to jointly optimize query relevance and user conversion objectives within a reinforcement learning paradigm. QueryAgent-R1 innovatively combines chain-of-retrieval optimization, memory-enhanced user profiling, and end-to-end alignment training. Extensive experiments on both industrial and public datasets demonstrate substantial improvements over strong baselines, with online A/B tests showing a 2.9% increase in click-through rate and a 3.1% gain in guided conversion rate.

Abstract

Query recommendation in e-commerce search aims to proactively suggest queries that match users' potential interests. However, existing methods mainly optimize query-level relevance, while neglecting whether the retrieved products align with users' downstream preferences. This mismatch often leads to high query click through rates (CTR) but low product conversion rates (CVR). To bridge this gap, we propose QueryAgent-R1, a memory-augmented agentic framework that improves end-to-end alignment via chain-of-retrieval optimization. Our QueryAgent-R1 grounds query generation in real inventory retrieval, allowing the agent to validate and refine queries based on retrieved products. We also design a consistency reward in the agentic reinforcement learning (RL) process to jointly optimize query relevance and downstream engagement. In addition, we construct a memory abstraction module for efficient user profiling. To support offline evaluation, we construct two datasets based on both proprietary industrial data and public datasets, on which QueryAgent-R1 consistently outperforms strong baselines. Moreover, on a large scale production platform, QueryAgent-R1 improves Query CTR by 2.9% and guided CVR by 3.1% in online A/B tests.

Problem

• query recommendation
• product retrieval
• conversion rate
• e-commerce search
• user preference alignment

Innovation

• query recommendation
• product retrieval
• agentic framework
• reinforcement learning
• memory augmentation