Table 1 Results of the credibility test

From: Element mining, network associations and scene reconstruction of qiantang river poetry road literary allusion landscapes

Evaluation dimensions

Allusion types

Sample size (items)

Precision

Recall

F1-Score

Classified by Allusion Types

Historical Events Category

185

0.91

0.89

0.9

Myths and Legends Category

128

0.88

0.86

0.87

Previous Generation Literature Category

109

0.85

0.79

0.82

Overall model performance

All types merged

422

0.89

0.86

0.87

Comparative experiment (baseline model)

BiLSTM+CRF

422

0.81

0.78

0.79

Special Evaluation of Low-Frequency Allusions

occurrences ≤ 5 times

63

0.80

0.75

0.77

  1. * (Drawn by the author).
Back to article page