An Automatic Error Detection Method for Engineering English Translation Based on the Deep Learning Model

Table 4

Experimental dataset table.

Experimental dataset number	Number of translated speeches	Proportion of error segments
01	53 paragraphs	5.61%
02	105 paragraphs	10.26%
03	205 paragraphs	15.82%
04	403 paragraphs	25.47%
05	804 paragraphs	30.13%
06	1605 paragraphs	35.47%
07	3206 paragraphs	38.78%
08	6408 paragraphs	40.13%
09	8002 paragraphs	41.59%
10	10005 paragraphs	46.21%