CNMBert: A Model For Hanyu Pinyin Abbreviation to Character Conversion Task

The task of converting Hanyu Pinyin abbreviations to Chinese characters represents a significant branch within the domain of Chinese Spelling Correction (CSC). This task is typically one of text-length alignment, however, due to the limited informational content in pinyin abbreviations, achieving accurate conversion is challenging. In this paper, we propose CNMBert which stands for zh-CN Pinyin Multi-mask Bert Model as a solution to this issue. CNMBert surpasses few-shot GPT models, achieving a 59.63% MRR on a 10,424-sample Hanyu Pinyin abbreviation test dataset.