Effects of Cross-lingual Evidence in Multilingual Medical Question Answering

2026-04-22

AI Summary

This study addresses the low accuracy of medical question answering in low-resource languages by systematically evaluating the impact of three types of external evidence: domain-specific medical knowledge bases, web-retrieved content, and parametric knowledge from large language models. It further compares monolingual, multilingual, and cross-lingual retrieval strategies. The findings reveal that external knowledge does not universally enhance performance; to address this, the authors propose an English–target language joint retrieval strategy that significantly improves answer accuracy for low-resource languages, bringing it close to the performance observed in high-resource settings. Experiments across language models of varying scales, integrating specialized resources like PubMed and general web retrieval, demonstrate the effectiveness of this approach: while large models excel on English tasks and high-resource languages benefit from English-only retrieval, low-resource languages achieve substantial gains through joint retrieval.

Technology Category

• Natural Language Processing: Question Answering
• Data Mining & Knowledge Management: Conversational Systems for Recommendation & Retrieval
• Machine Learning: Large Multimodal Models (LMMs)

Application Category

• Search and Retrieval-Augmented AI: Multilingual and cross-lingual Web search
• Web Mining and Content Analysis: Mining multimedia, multimodal, multilingual, cross-lingual Web data
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

This paper investigates Multilingual Medical Question Answering across high-resource (English, Spanish, French, Italian) and low-resource (Basque, Kazakh) languages. We evaluate three types of external evidence sources across models of varying size: curated repositories of specialized medical knowledge, web-retrieved content, and explanations from LLM's parametric knowledge. Moreover, we conduct experiments with multilingual, monolingual and cross-lingual retrieval. Our results demonstrate that larger models consistently achieve superior performance in English across baseline evaluations. When incorporating external knowledge, web-retrieved data in English proves most beneficial for high-resource languages. Conversely, for low-resource languages, the most effective strategy combines retrieval in both English and the target language, achieving comparable accuracy to high-resource language results. These findings challenge the assumption that external knowledge systematically improves performance and reveal that effective strategies depend on both the source of language resources and on model scale. Furthermore, specialized medical knowledge sources such as PubMed are limited: while they provide authoritative expert knowledge, they lack adequate multilingual coverage

Problem

• Multilingual Medical Question Answering
• Cross-lingual Evidence
• Low-resource Languages
• External Knowledge
• Language Resource Coverage

Innovation

• Multilingual Medical Question Answering
• Cross-lingual Retrieval
• Low-resource Languages
• External Knowledge Integration
• Model Scale