Which Encoding is the Best for Text Classification in Chinese, English, Japanese and Korean?
This work addresses encoding challenges for multilingual text classification, providing practical insights but is incremental as it compares existing methods without introducing new ones.
The paper empirically studied various encoding methods for text classification across Chinese, English, Japanese, and Korean, evaluating 473 models on 14 datasets, finding that UTF-8 byte-level one-hot encoding is competitive for convolutional networks and word-level n-grams work well without perfect segmentation.
This article offers an empirical study on the different ways of encoding Chinese, Japanese, Korean (CJK) and English languages for text classification. Different encoding levels are studied, including UTF-8 bytes, characters, words, romanized characters and romanized words. For all encoding levels, whenever applicable, we provide comparisons with linear models, fastText and convolutional networks. For convolutional networks, we compare between encoding mechanisms using character glyph images, one-hot (or one-of-n) encoding, and embedding. In total there are 473 models, using 14 large-scale text classification datasets in 4 languages including Chinese, English, Japanese and Korean. Some conclusions from these results include that byte-level one-hot encoding based on UTF-8 consistently produces competitive results for convolutional networks, that word-level n-grams linear models are competitive even without perfect word segmentation, and that fastText provides the best result using character-level n-gram encoding but can overfit when the features are overly rich.