This research introduces a novel explainable LLM-driven framework for e-commerce search that improves query-item relevance while maintaining efficiency.

• Leverages LLMs' knowledge while addressing their practical limitations in e-commerce environments
• Implements multi-dimensional distillation to transfer LLM capabilities to smaller, specialized models
• Creates explainable relevance judgments through semantic dimension modeling
• Achieves superior performance particularly for long-tail queries

This engineering advancement matters because it bridges the gap between powerful but resource-intensive LLMs and the practical needs of e-commerce platforms, enabling better search experiences without compromising system efficiency.

Explainable LLM-driven Multi-dimensional Distillation for E-Commerce Relevance Learning