From Unstructured to Structured: LLM-Guided Attribute Graphs for Entity Search and Ranking
Yilun Zhu, Nikhita Vedula, Shervin Malmasi
TLDR
This paper introduces an LLM-driven two-stage approach using attribute graphs for improved entity search and ranking in e-commerce, boosting precision.
Key contributions
- Proposes a two-stage LLM-driven approach for entity search: attribute graph construction and graph-aware ranking.
- Extracts structured product attributes from text to build category-aware, reusable attribute graphs offline.
- Ranks candidates online using structured representations, reducing token usage by 57% and improving precision.
- Achieves over 5% improvement in average precision over baselines in zero-shot scenarios without training data.
Why it matters
This paper addresses the challenge of nuanced entity search in e-commerce, where traditional methods fall short. By leveraging LLMs to create structured attribute graphs, it significantly improves ranking precision and efficiency. This approach offers robust, zero-shot performance, making it highly practical for real-world deployment.
Original Abstract
Entity search, i.e., finding the most similar entities to a query entity, faces unique challenges in e-commerce, where product similarity varies across categories and contexts. Traditional embedding-based approaches often struggle to capture nuanced context-specific attribute relevance. In this paper, we present a two-stage approach combining Large Language Model (LLM)-driven attribute graph construction with graph-aware LLM ranking. In the offline stage, we extract structured product attributes from unstructured text, and construct a reusable attribute graph with category-aware schemas. In the online stage, we rank retrieved candidates by reasoning over this structured representation rather than raw text, reducing per-product token usage by 57% while improving ranking precision. Experiments show that our approach outperforms multiple baselines under zero-shot scenarios, achieving a over 5% improvement in average precision without requiring training data, generalizes robustly across diverse product categories, and shows immense potential for real-world deployment.
📬 Weekly AI Paper Digest
Get the top 10 AI/ML arXiv papers from the week — summarized, scored, and delivered to your inbox every Monday.