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

Table 10

Result example of confusion matrix for entity extraction.

True/predict	Date	Dept	Work	Name	Time	Num	No_Entity
Date	100	0	0	0	0	0	0
Dept	0	242	0	0	0	4	0
Work	4	4	567	2	0	0	63
Name	0	0	20	29	0	0	5
Time	0	0	0	0	13	0	0
Num	5	0	0	0	0	15	0
No_Entity	0	2	23	3	0	0	2253