Contextual Embedding Replacement vs. Bilingual Lexicon Substitution for Zero-Shot Cross-Lingual Retrieval Robustness

Transferring information retrieval (IR) models from a high-resource language (typically English) to other languages in a zero-shot fashion has become a widely adopted approach. In this work, we show that the effectiveness of zero-shot rankers diminishes when queries and documents are present in different languages. Motivated by this, we propose to train ranking models on artificially code-switched data instead, which we generate by utilizing bilingual lexicons. To this end, we experiment with lexicons induced from (1) cross-lingual word embeddings and (2) parallel Wikipedia page titles. We use

Research goal: How does training on artificially code-switched data generated via contextual embedding replacement compare to bilingual lexicon substitution for improving zero-shot cross-lingual retrieval robustness on the MIRACL benchmark?

Autonomous synthesis report generated by Assignee Research. Tribunal consensus score: 8.2/10.