[Retracted] Design of Automatic Scoring System for Oral English Test Based on Sequence Matching and Big Data Analysis

Table 2

Average difference and consistency results of machine evaluation and human evaluation.

Types of oral test questions	Human evaluation and machine evaluation	Dialogue	Description	Impromptu composition	Recite
Complete consistency between human evaluation and machine evaluation/%	and U1	0.34	21.24	12.23	34.71
	and U2	43.91	3.49	3.14	81.31
	and U3	43.55	2.41	43.01	21.08
	and U4	28.71	56.23	86.26	19.75
Proximity consistency between human evaluation and machine evaluation/%	and U1	14.91	74.24	73.45	29.12
	and U2	24.34	50.12	81.34	43.42
	and U3	81.4	38.23	48.01	34.11
	and U4	4.12	47.13	74.85	42.22
Correlation coefficient between human evaluation and machine evaluation/%	and U1	65.78	29.48	14.47	21.89
	and U2	85.67	92.9	74.17	21.87
	and U3	14.1	18.75	65.32	30.24
	and U4	40.11	31.77	39.09	42.87

.