Improving Korean NLP Tasks with Linguistically Informed Subword Tokenization and Sub-character Decomposition

Open in new window