English

A Chat About Boring Problems: Studying GPT-based text normalization

Computation and Language 2024-01-18 v2 Artificial Intelligence

Abstract

Text normalization - the conversion of text from written to spoken form - is traditionally assumed to be an ill-formed task for language models. In this work, we argue otherwise. We empirically show the capacity of Large-Language Models (LLM) for text normalization in few-shot scenarios. Combining self-consistency reasoning with linguistic-informed prompt engineering, we find LLM based text normalization to achieve error rates around 40\% lower than top normalization systems. Further, upon error analysis, we note key limitations in the conventional design of text normalization tasks. We create a new taxonomy of text normalization errors and apply it to results from GPT-3.5-Turbo and GPT-4.0. Through this new framework, we can identify strengths and weaknesses of GPT-based TN, opening opportunities for future work.

Keywords

Cite

@article{arxiv.2309.13426,
  title  = {A Chat About Boring Problems: Studying GPT-based text normalization},
  author = {Yang Zhang and Travis M. Bartley and Mariana Graterol-Fuenmayor and Vitaly Lavrukhin and Evelina Bakhturina and Boris Ginsburg},
  journal= {arXiv preprint arXiv:2309.13426},
  year   = {2024}
}

Comments

Accepted to ICASSP 2024

R2 v1 2026-06-28T12:30:29.753Z