@inproceedings{liang-etal-2025-colloquial,
    title = "Colloquial Singaporean {E}nglish Style Transfer with Fine-Grained Explainable Control",
    author = "Liang, Jinggui  and
      Vo, Dung  and
      Xian, Yap Hong  and
      Chieu, Hai Leong  and
      Chai, Kian Ming A.  and
      Jiang, Jing  and
      Liao, Lizi",
    editor = "Che, Wanxiang  and
      Nabende, Joyce  and
      Shutova, Ekaterina  and
      Pilehvar, Mohammad Taher",
    booktitle = "Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)",
    month = jul,
    year = "2025",
    address = "Vienna, Austria",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/ingestion-acl-25/2025.acl-long.1309/",
    pages = "26962--26983",
    ISBN = "979-8-89176-251-0",
    abstract = "Colloquial Singaporean English (Singlish) is an informal English marked by a unique blend of languages reflecting Singapore{'}s multicultural identity. Style transfer between Singlish and Standard (formal) English is vital for various applications, yet existing methods often lack explainability and fine-grained control. To fill this gap, we contribute in two key ways. First, we construct a large, high-quality dataset of formal and informal sentences, annotated across six linguistic aspects{---}Syntax, Lexical Borrowing, Pragmatics, Prosody/Phonology, Emoticons/Punctuation, and Code-Switching{---}with detailed explanations. Starting with manually annotated cases, we scaled the dataset to 140K with ensured quality. Second, inspired by the ``Society of Mind'' theory, we propose a novel multi-agent framework where large language models (LLMs) act as expert agents for each linguistic aspect. These agents collaborate by iteratively generating, critiquing, and refining responses to achieve controlled, explainable style transfer. Both automatic metrics and human evaluations confirm that our method enables precise, interpretable transformations, advancing explainability in NLP for Singlish."
}