An Empirical Study of Tokenization Strategies for Various Korean NLP Tasks

454

Knowledge Triples

Source Entity	Relationship	Target Entity
Machine Translation Task	uses	KorNLI Material
Machine Translation Task	uses	KorQuAD Material
Machine Translation Task	uses	KorSTS Material
Machine Translation Task	uses	NSMC Material
Machine Translation Task	uses	PAWS X Material
Natural Language Understanding Task	uses	KorNLI Material
Natural Language Understanding Task	uses	KorQuAD Material
Natural Language Understanding Task	uses	KorSTS Material
Natural Language Understanding Task	uses	NSMC Material
Natural Language Understanding Task	uses	PAWS X Material
Tokenization Method	appliesTo	Machine Translation Task
Tokenization Method	appliesTo	Natural Language Understanding Task