Eager improves e-commerce search query suggestions using clicked items
EAGER: Enrich-and-Align Generative Query Recommendation from Clicked Items in E-commerce Search
Information Retrieval
Summary
Online shopping sites often show suggested searches to help users find what they want more easily. The authors found that previous methods either relied too much on old data or just generated generic suggestions not based on real user behavior. They developed EAGER, a two-step method that learns from the items users clicked on to create better, more personalized search suggestions. This system was tested extensively and is now used in a major online shopping platform.
What this means in practice
- •For e-commerce platform engineers: Generate personalized search query suggestions based on user clicks to improve search experience and engagement.$Commercial implications: Enables selling a tailored search recommendation feature to online stores for better user retention and sales.
- •For digital marketing teams: Use enriched click-based query suggestions to design more effective search campaigns targeting user intent.
Authors
Shuwei Yuan, Mingqian Ding, Luxin Liu, Rong Xiao, Xiaoyi Zeng
Abstract
E-commerce platforms increasingly display clickable query suggestions alongside items in the user feed, enabling users to refine or expand their intent without manually reformulating queries. Existing approaches either mine suggestions from historical logs -- limited to past behavior and blind to long-tail, personalized intents -- or rely on off-the-shelf LLMs whose lack of platform-specific knowledge yields fluent but generic queries disconnected from real click behavior. We propose EAGER (Enrich-and-AliGn gEnerative Query Recommendation), a two-stage framework for generating query suggestions from clicked items. In the enrichment stage, supervised fine-tuning (SFT) follows a four-stage curriculum that scales information richness (from item-only to user-conditioned) and reasoning depth (from direct to chain-of-thought). Each stage incorporates rationale augmentation, diversity regularization, and self-distillation. In the alignment stage, we post-train via GRPO with a hybrid reward of multiple rule-based business signals and a preference-aware click reward. Extensive offline experiments and online A/B test demonstrate the effectiveness of EAGER, which has been deployed in production at a major e-commerce platform.