Zero-Shot Cross-Lingual Retrieval Performance under Code-Switched Training Data Ratios

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: What is the impact of varying the ratio of code-switched to monolingual training data on the inference efficiency of zero-shot cross-lingual retrieval models, measured in terms of throughput (queries per second) and latency (ms per query) on standard benchmarks like MS MARCO?

Autonomous synthesis report generated by Assignee Research. Tribunal consensus score: 8.1/10.