What Drives Cross-lingual Ranking? Retrieval Approaches with Multilingual Language Models

TL;DR

This paper evaluates various retrieval approaches using multilingual models for cross-lingual information retrieval, highlighting the effectiveness of dense retrieval and contrastive learning over translation-based methods, especially for low-resource languages.

Contribution

It systematically compares four intervention types for CLIR, demonstrating the superiority of dense retrieval and contrastive learning in multilingual settings.

Findings

Dense retrieval models outperform lexical matching and translation-based methods.

Contrastive learning improves alignment, especially for weakly aligned encoders.

Re-ranking effectiveness depends on training data quality.

Abstract

Cross-lingual information retrieval (CLIR) enables access to multilingual knowledge but remains challenging due to disparities in resources, scripts, and weak cross-lingual semantic alignment in embedding models. Existing pipelines often rely on translation and monolingual retrieval heuristics, which add computational overhead and noise, degrading performance. This work systematically evaluates four intervention types, namely document translation, multilingual dense retrieval with pretrained encoders, contrastive learning at word, phrase, and query-document levels, and cross-encoder re-ranking, across three benchmark datasets. We find that dense retrieval models trained specifically for CLIR consistently outperform lexical matching methods and derive little benefit from document translation. Contrastive learning mitigates language biases and yields substantial improvements for encoders with weak initial alignment, and re-ranking can be effective, but depends on the quality of the cross-encoder training data. Although high-resource languages still dominate overall performance, gains over lexical and document-translated baselines are most pronounced for low-resource and cross-script pairs. These findings indicate that cross-lingual search systems should prioritise semantic multilingual embeddings and targeted learning-based alignment over translation-based pipelines, particularly for cross-script and under-resourced languages.