Know When to Fuse: Investigating Non-English Hybrid Retrieval in the Legal Domain

TL;DR

This study explores hybrid retrieval methods in French legal texts, revealing that fusion improves zero-shot performance but may reduce in-domain results unless carefully tuned, thus expanding understanding of non-English domain-specific search.

Contribution

It investigates hybrid search effectiveness in French legal retrieval, highlighting differences between zero-shot and in-domain scenarios and providing new insights into fusion strategies in non-English contexts.

Findings

Fusion improves zero-shot retrieval performance.

In-domain fusion often reduces accuracy unless scores are carefully weighted.

Study extends hybrid search understanding to non-English, specialized domains.

Abstract

Hybrid search has emerged as an effective strategy to offset the limitations of different matching paradigms, especially in out-of-domain contexts where notable improvements in retrieval quality have been observed. However, existing research predominantly focuses on a limited set of retrieval methods, evaluated in pairs on domain-general datasets exclusively in English. In this work, we study the efficacy of hybrid search across a variety of prominent retrieval models within the unexplored field of law in the French language, assessing both zero-shot and in-domain scenarios. Our findings reveal that in a zero-shot context, fusing different domain-general models consistently enhances performance compared to using a standalone model, regardless of the fusion method. Surprisingly, when models are trained in-domain, we find that fusion generally diminishes performance relative to using the best single system, unless fusing scores with carefully tuned weights. These novel insights, among others, expand the applicability of prior findings across a new field and language, and contribute to a deeper understanding of hybrid search in non-English specialized domains.