Table 12 Bi-lstm model results analysis of each feaure (%).

From: An AI-driven tools assessment framework for english teachers using the Fuzzy Delphi algorithm and deep learning

Features

Accuracy

Precision

Recall

F1-Score

Learning Enhancement

77

73

75

70

Engagement

83

86

89

81

Skill Development

90

92

98

97

Clarity Cohesion

88

89

80

81

Analytical Quality

90

91

93

97

Usability

91

89

87

91

Overall Model Performance

90

92

93

92