Enregistré dans:
Détails bibliographiques
Auteurs principaux: Lai, Foong Ming, Tan, Yujin, Meng, Han, Lee, Yi-Chieh
Format: Preprint
Publié: 2026
Sujets:
Accès en ligne:https://arxiv.org/abs/2605.07132
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866917471346229248
author Lai, Foong Ming
Tan, Yujin
Meng, Han
Lee, Yi-Chieh
author_facet Lai, Foong Ming
Tan, Yujin
Meng, Han
Lee, Yi-Chieh
contents Code-switching in contact varieties like Singaporean English (Singlish) challenges natural language generation due to limited parallel data and rapid lexical evolution. We propose a retrieval-augmented generation (RAG) framework that externalizes dialectal knowledge into a curated lexicon, enabling controlled lexical code-switching without fine-tuning. Our approach retrieves candidate Singlish expressions and guides generation through sparse lexical substitution. Human evaluation with 164 Singaporean participants found RAG and zero-shot prompting equally natural and appropriate. Automatic analyses reveal different transformation regimes: zero-shot prompting induces extensive paraphrasing (median 23 token edits), whereas RAG performs minimal substitutions (median 1 edit) with higher semantic preservation (mean cosine similarity 0.978 vs. 0.926). Our results demonstrate that externalizing code-switching into lexical resources enables control and auditability without sacrificing perceived quality, offering practical advantages for rapidly evolving contact varieties.
format Preprint
id arxiv_https___arxiv_org_abs_2605_07132
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle From Standard English to Singlish: A Retrieval-Augmented Approach for Code-Switched Creole Generation in Large Language Models
Lai, Foong Ming
Tan, Yujin
Meng, Han
Lee, Yi-Chieh
Human-Computer Interaction
Code-switching in contact varieties like Singaporean English (Singlish) challenges natural language generation due to limited parallel data and rapid lexical evolution. We propose a retrieval-augmented generation (RAG) framework that externalizes dialectal knowledge into a curated lexicon, enabling controlled lexical code-switching without fine-tuning. Our approach retrieves candidate Singlish expressions and guides generation through sparse lexical substitution. Human evaluation with 164 Singaporean participants found RAG and zero-shot prompting equally natural and appropriate. Automatic analyses reveal different transformation regimes: zero-shot prompting induces extensive paraphrasing (median 23 token edits), whereas RAG performs minimal substitutions (median 1 edit) with higher semantic preservation (mean cosine similarity 0.978 vs. 0.926). Our results demonstrate that externalizing code-switching into lexical resources enables control and auditability without sacrificing perceived quality, offering practical advantages for rapidly evolving contact varieties.
title From Standard English to Singlish: A Retrieval-Augmented Approach for Code-Switched Creole Generation in Large Language Models
topic Human-Computer Interaction
url https://arxiv.org/abs/2605.07132