Enhancement of E-commerce Sponsored Search Relevancy with LLM

2026-07-04

AI Summary

This study addresses the challenges of relevance matching in e-commerce sponsored search, where the vast keyword space, ambiguous user intent, and multilingual, multi-topic nature of queries complicate accurate ad-item alignment. To tackle this, the authors propose a domain-adapted relevance classification method based on the LLaMA2-7B large language model. By leveraging Low-Rank Adaptation (LoRA) for efficient fine-tuning, they apply LLaMA2 to the e-commerce advertising context for the first time while maintaining privacy and manageable inference costs. The approach introduces a fine-grained three-class relevance judgment—relevant, partially relevant, and irrelevant. Experimental results on a large-scale test set demonstrate an accuracy of 89.43%, significantly outperforming existing baseline models and even advanced language models such as GPT-4.

Abstract

Sponsored search plays a crucial role as a revenue stream for search engines, wherein advertisers competitively bid on keywords that align with the users' search queries. The task of matching relevant keywords to these queries is complicated by the vast and ever-evolving space of keywords, the ambiguity of user and advertiser intentions, and the wide range of topics and languages involved. Consequently, ensuring that ads are pertinent to user queries presents significant challenges. In the fast-paced world of e-commerce, the accuracy of sponsored search results is vital for boosting user satisfaction and optimizing business operations. This paper presents the development of an advanced Ad Relevance Model within a sponsored search framework, utilizing the power of a pretrained large language model. We detail a pioneering adaptation of the LLAMA2 7B model through Low-Rank Adaptation (LoRA), which markedly enhances search precision and operational efficiency, thus opening new avenues for improving user interactions in extensive online marketplaces such as Walmart.com. We introduce a novel query and ad title classifier, which discerns the relevance of search interactions across three categories: Relevant, Partially Relevant, and Irrelevant. Our approach involved adapting the pretrained model specifically for the e-commerce sponsored search context, training it on a large dataset. The fine-tuned model demonstrated a marked improvement in ad relevance accuracy, achieving 89.43% accuracy on a comprehensive test dataset -- outperforming both the baseline model and other advanced language models like GPT-4. The integration of LoRA with the based model represents a significant stride in customizing language models for e-commerce applications, resulting in enhanced search accuracy, cost efficiency, and operational privacy -- a triad essential for the modern digital marketplace.

Problem

• sponsored search
• ad relevance
• e-commerce
• keyword-query matching
• user intent ambiguity

Innovation

• Large Language Model
• LoRA
• Sponsored Search
• Ad Relevance
• E-commerce