Semantic Ads Retrieval at Walmart eCommerce with Language Models Progressively Trained on Multiple Knowledge Domains

2025-02-13

AI Summary

E-commerce search advertising faces core challenges including linguistic asymmetry between queries and product titles, ambiguous user intent, and imbalanced, sparse training corpora. Method: This paper develops an end-to-end semantic retrieval system for Walmart, introducing a human-feedback-driven, progressive multi-domain knowledge fusion training paradigm. It pioneers the joint optimization of category-aware BERT pretraining and a dual-tower Siamese architecture. Contribution/Results: The approach achieves significant gains in semantic matching accuracy while maintaining high engineering efficiency. Compared to baseline models, search relevance improves by 16%; large-scale online A/B tests confirm substantial increases in advertising revenue. The system comprehensively outperforms Walmart’s current production system across all key metrics, demonstrating robust generalization and operational scalability.

Technology Category

• Search and Optimization: Learning to Search
• Data Mining & Knowledge Management: Conversational Systems for Recommendation & Retrieval
• Constraint Satisfaction and Optimization: Search

Application Category

• Search and Retrieval-Augmented AI: Ad search and search for Web retail
• Semantics and Knowledge: Applications of semantic technologies for improving search, browsing, recommendation, personalization
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Sponsored search in e-commerce poses several unique and complex challenges. These challenges stem from factors such as the asymmetric language structure between search queries and product names, the inherent ambiguity in user search intent, and the vast volume of sparse and imbalanced search corpus data. The role of the retrieval component within a sponsored search system is pivotal, serving as the initial step that directly affects the subsequent ranking and bidding systems. In this paper, we present an end-to-end solution tailored to optimize the ads retrieval system on Walmart.com. Our approach is to pretrain the BERT-like classification model with product category information, enhancing the model's understanding of Walmart product semantics. Second, we design a two-tower Siamese Network structure for embedding structures to augment training efficiency. Third, we introduce a Human-in-the-loop Progressive Fusion Training method to ensure robust model performance. Our results demonstrate the effectiveness of this pipeline. It enhances the search relevance metric by up to 16% compared to a baseline DSSM-based model. Moreover, our large-scale online A/B testing demonstrates that our approach surpasses the ad revenue of the existing production model.

Problem

• Optimize semantic ads retrieval in e-commerce
• Handle asymmetric language and ambiguous intent
• Improve search relevance and ad revenue

Innovation

• BERT-like model pretraining
• Siamese Network structure
• Human-in-the-loop training