Fig. 4From: Korean clinical entity recognition from diagnosis text using BERTAn example of tokenized sentence in diagnosis dataset. Each column corresponds to the output labels, input tokens, and the translation of input in EnglishBack to article page