Shared Heritage, Distinct Writing: Rethinking Resource Selection for East Asian Historical Documents
This challenges a common practice in low-resource historical document processing, emphasizing the need for empirical validation rather than assumed cross-lingual transfer.
The paper tackled the assumption that Classical Chinese resources improve processing of historical Korean and Japanese documents, finding minimal impact with performance differences within ±0.0068 F1-score for sequence labeling and up to +0.84 BLEU for translation, and that benefits diminish as local data increases.
Historical documents in the Sinosphere are known to share common formats and practices, particularly in veritable records compiled by court historians. This shared linguistic heritage has led researchers to use Classical Chinese resources for cross-lingual transfer when processing historical documents from Korea and Japan, which remain relatively low-resource. In this paper, we question the assumption of cross-lingual transferability from Classical Chinese to Hanja and Kanbun, the ancient written languages of Korea and Japan, respectively. Our experiments across machine translation, named entity recognition, and punctuation restoration tasks show minimal impact of Classical Chinese datasets on language model performance for ancient Korean documents written in Hanja, with performance differences within $\pm{}0.0068$ F1-score for sequence labeling tasks and up to $+0.84$ BLEU score for translation. These limitations persist consistently across various model sizes, architectures, and domain-specific datasets. Our analysis reveals that the benefits of Classical Chinese resources diminish rapidly as local language data increases for Hanja, while showing substantial improvements only in extremely low-resource scenarios for both Korean and Japanese historical documents. These findings emphasize the need for careful empirical validation rather than assuming benefits from indiscriminate cross-lingual transfer.