Paper Detail
Qingcheng Zeng, Yuheng Lu, Zeqi Zhou, Heli Qi, Puxuan Yu, Fuheng Zhao, Hitomi Yanaka, Weihao Xuan, Naoto Yokoya
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.
No structured notes yet. Add `summary_sections`, `why_relevant`, `claim_impact`, or `next_action` in `papers.jsonl` to enrich this view.
No ranking explanation is available yet.
No tags.
@misc{zeng2026code,
title = {Code-Switching Information Retrieval: Benchmarks, Analysis, and the Limits of Current Retrievers},
author = {Qingcheng Zeng and Yuheng Lu and Zeqi Zhou and Heli Qi and Puxuan Yu and Fuheng Zhao and Hitomi Yanaka and Weihao Xuan and Naoto Yokoya},
year = {2026},
abstract = {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, de},
url = {https://huggingface.co/papers/2604.17632},
keywords = {code-switching, information retrieval, multilingual models, embedding space, vocabulary expansion, CSR-L, CS-MTEB, code available, huggingface daily},
eprint = {2604.17632},
archiveprefix = {arXiv},
}
{}