RRADistill: Distilling LLMs' Passage Ranking Ability for Long-Tail Queries Document Re-Ranking on a Search Engine

TL;DR

This paper introduces RRADistill, a method to efficiently distill LLMs' passage ranking skills into smaller models, enhancing long-tail query re-ranking in search engines with validated improvements through A/B testing.

Contribution

The paper presents a novel distillation pipeline and training methods for small LLMs, leveraging LLM capabilities more effectively for passage re-ranking tasks.

Findings

Improved re-ranking accuracy for long-tail queries.

Efficient label generation pipeline for distillation.

Validated effectiveness through A/B testing on a real search platform.

Abstract

Large Language Models (LLMs) excel at understanding the semantic relationships between queries and documents, even with lengthy and complex long-tail queries. These queries are challenging for feedback-based rankings due to sparse user engagement and limited feedback, making LLMs' ranking ability highly valuable. However, the large size and slow inference of LLMs necessitate the development of smaller, more efficient models (sLLMs). Recently, integrating ranking label generation into distillation techniques has become crucial, but existing methods underutilize LLMs' capabilities and are cumbersome. Our research, RRADistill: Re-Ranking Ability Distillation, propose an efficient label generation pipeline and novel sLLM training methods for both encoder and decoder models. We introduce an encoder-based method using a Term Control Layer to capture term matching signals and a decoder-based model with a ranking layer for enhanced understanding. A/B testing on a Korean-based search platform, validates the effectiveness of our approach in improving re-ranking for long-tail queries.