Automatic Extraction of English-Chinese Translation Templates Based on Deep Learning

Table 3

Performance test results of the ATTEBSC algorithm.

	Corpus scale	Corpus similarity	Running time (seconds)	A template for	Number of dictionary	Dead template ratio (%)
S1	4000	Poor	141.19	10921	9294	19
	10000	Poor	1819.14	98928	88291	81
S2	4000	Poor	118.92	914	901	49
	10000	Poor	1094.44	8820	1218	48