Investigating Glyph Phonetic Information for Chinese Spell Checking: What Works and What's Next

TL;DR

This paper examines how glyph and phonetic information influence Chinese spell checking models, assesses their generalization capabilities, and proposes a new challenging testing setting to guide future improvements.

Contribution

It provides insights into the use of glyph-phonetic information in CSC and introduces a more rigorous evaluation framework for model generalization.

Findings

Glyph-phonetic information improves spell checking accuracy.

Current models have limited generalization to new error types.

Proposed setting reveals gaps in model robustness.

Abstract

While pre-trained Chinese language models have demonstrated impressive performance on a wide range of NLP tasks, the Chinese Spell Checking (CSC) task remains a challenge. Previous research has explored using information such as glyphs and phonetics to improve the ability to distinguish misspelled characters, with good results. However, the generalization ability of these models is not well understood: it is unclear whether they incorporate glyph-phonetic information and, if so, whether this information is fully utilized. In this paper, we aim to better understand the role of glyph-phonetic information in the CSC task and suggest directions for improvement. Additionally, we propose a new, more challenging, and practical setting for testing the generalizability of CSC models. All code is made publicly available.