Accès libre

Intelligent Construction of Civic Teaching Resources for Ancient Literature Course Based on Natural Language Processing Technology

  
21 mars 2025
À propos de cet article

Citez
Télécharger la couverture

Figure 1.

BERT model structure
BERT model structure

Figure 2.

BERT structure after improvement
BERT structure after improvement

Figure 3.

TF-IDF score
TF-IDF score

Figure 4.

Comparison of comprehensive scores
Comparison of comprehensive scores

Figure 5.

P/N measure results
P/N measure results

Figure 6.

AUC test results of different methods
AUC test results of different methods

The coverage test results of different methods

Epoch Coverage rate/% Epoch Coverage rate/%
Ours Method1 Method2 Ours Method1 Method2
1 99.13 75.36 59.42 11 99.05 74.83 61.57
2 98.62 77.26 64.22 12 99.80 76.15 58.38
3 98.66 75.69 57.43 13 98.32 76.08 62.32
4 98.71 75.03 57.45 14 99.5 74.75 57.00
5 98.75 73.65 60.45 15 98.22 73.95 64.33
6 98.23 76.55 60.14 16 98.37 76.14 60.35
7 98.82 73.90 65.48 17 99.24 75.14 60.78
8 99.29 74.62 60.17 18 99.50 73.09 59.23
9 98.89 74.31 57.71 19 99.59 75.05 62.34
10 98.53 72.85 59.62 20 99.28 75.55 63.75

Knowledge matching matrix

Knowledge item K1 K2 Kn
K1 M11 M12 M1n
K2 M21 M22 M2n
Kn Mn1 Mn2 Mnn

Knowledge relevance matrix

Knowledge item K1 K2 Kn
K1 asso11 asso12 asso1n
K2 asso21 asso22 asso2n
Kn asson1 asson2 assonn

Knowledge similarity matrix

Knowledge item K1 K2 Kn
K1 Sim11 Sim12 Sim1n
K2 Sim21 Sim22 Sim2n
Kn Simn1 Simn2 Simnn