Code-Switching Information Retrieval: Benchmarks, Analysis, and the Limits of Current Retrievers
For the information retrieval community, this work highlights code-switching as a critical performance bottleneck that current systems fail to handle, establishing a new frontier for optimization.
This paper introduces CSR-L and CS-MTEB benchmarks for code-switching information retrieval, finding that code-switching degrades retriever performance by up to 27% and that standard multilingual techniques are insufficient to address the issue.
Code-switching is a pervasive linguistic phenomenon in global communication, yet modern information retrieval systems remain predominantly designed for, and evaluated within, monolingual contexts. To bridge this critical disconnect, we present a holistic study dedicated to code-switching IR. We introduce CSR-L (Code-Switching Retrieval benchmark-Lite), constructing a dataset via human annotation to capture the authentic naturalness of mixed-language queries. Our evaluation across statistical, dense, and late-interaction paradigms reveals that code-switching acts as a fundamental performance bottleneck, degrading the effectiveness of even robust multilingual models. We demonstrate that this failure stems from substantial divergence in the embedding space between pure and code-switched text. Scaling this investigation, we propose CS-MTEB, a comprehensive benchmark covering 11 diverse tasks, where we observe performance declines of up to 27%. Finally, we show that standard multilingual techniques like vocabulary expansion are insufficient to resolve these deficits completely. These findings underscore the fragility of current systems and establish code-switching as a crucial frontier for future IR optimization.