Human Evaluation Results

Table 4 summarizes human evaluation results of our machine translation evaluation for the JE and EJ tasks. In this table, ``Proposed'' indicates our proposed system. ``adequacy'' indicates the average adequacy . ``acceptability'' indicates the average acceptability. ``()'' means the order of all entry systems. For example, Our system was the 7th place in 19 system for the average adequacy in JE task.

表 4: Results of Human Evaluation

	task	adequacy	acceptability
			pairwise comparison score	(tie)
Proposed	JE	2.73	0.4604	0.3312
(RBMT+SMT)		(7/19)	(8/14)	(9/14)
Proposed	EJ	2.6	0.4318	0.2992
(RBMT+SMT)		(9/17)	(8/11)	(5/11)

As seen in these results, our method was so effective. And the BLUE sore was worse compared to other systems. Howver results of human evaluation was good compared to other systems.