KoRASA: Pipeline Optimization for Open-Source Korean Natural Language Understanding Framework Based on Deep Learning

Table 5

Tokenization.

Model	Intent classification				Entity extraction
	Train		Test		Train		Test
	Acc	F1	Acc	F1	Acc	F1	Acc	F1
WhiteSpace	1.000	1.000	0.984	0.981	0.085	0.073	0.073	0.050
ConveRT	1.000	1.000	0.982	0.978	0.085	0.073	0.071	0.047
Mecab (Ko)	1.000	1.000	0.980	0.985	0.993	0.989	0.959	0.905