Acceso abierto

Research on the Innovative Model of English Teaching by Integrating Traditional Culture and Artificial Intelligence

  
29 sept 2025

Cite
Descargar portada

Figure 1.

DKVMN-F Model Framework
DKVMN-F Model Framework

Figure 2.

Flowchart of the HANN framework
Flowchart of the HANN framework

Figure 3.

Comparison results of knowledge states of different models
Comparison results of knowledge states of different models

Figure 4.

ASSISTments2009 Data clustering class
ASSISTments2009 Data clustering class

Figure 5.

Influence of different comparative loss weights on Precision@5 and Recall@5
Influence of different comparative loss weights on Precision@5 and Recall@5

Figure 6.

Online personalized recommendation strategy
Online personalized recommendation strategy

AUC values of DKVMN-F variants on different datasets

Model ASSISTments2009 Statics2011 ASSISTments2012 Synthetic-5
DKVMN-F (basic) 0.8288 0.8386 0.7412 0.8406
DKVMN-F (without forget) 0.8481 0.8586 0.7560 0.8610
DKVMN-F (without learning) 0.8463 0.8587 0.7519 0.8567
DKVMN-F 0.8669 0.8670 0.7757 0.8701

Comparison of model performance on ASSISTments2009

Metric MAP Precision@1 Precision@5 Precision@10 Recall@1 Recall@5 Recall@10
Model
Caser 0.0303 0.0490 0.0451 0.0432 0.0028 0.0135 0.0249
RCNN 0.0327 0.0489 0.0481 0.0491 0.0022 0.0131 0.0269
CosRec 0.0375 0.0591 0.0544 0.0515 0.0038 0.0177 0.0354
SCosRec 0.0394 0.0604 0.0572 0.0537 0.0036 0.0211 0.0367
HANN 0.0431 0.0661 0.0614 0.0565 0.0041 0.0218 0.0389

Comparison of AUC on different data sets

Model ASSISTments2009 Statics2011 ASSISTments2012 Synthetic-5
DKT 0.7437 0.8142 0.7197 0.7598
DKT+forget 0.7528 0.7540 0.7354 0.7621
DKVMN 0.8254 0.8395 0.7362 0.8370
LFKT 0.8569 0.8619 0.7605 0.8614
Bi-CLKT 0.8549 0.8628 0.7617 0.8646
DKVMN-F 0.8669 0.8670 0.7757 0.8701

Comparison of model performance on ML-1M

Metric MAP Precision@1 Precision@5 Precision@10 Recall@1 Recall@5 Recall@10
Model
Pop 0.0694 0.1279 0.1122 0.1009 0.0055 0.0225 0.0369
BPR 0.0914 0.1472 0.1292 0.1183 0.0067 0.0301 0.0564
FPMC 0.1033 0.2004 0.1674 0.1448 0.0132 0.0454 0.0772
GRU4rec 0.1435 0.2512 0.2135 0.1914 0.0158 0.0618 0.1094
Caser 0.1512 0.2501 0.2189 0.1994 0.0146 0.0634 0.1118
RCNN 0.1681 0.2830 0.2491 0.2225 0.0191 0.0728 0.1267
CosRec 0.1895 0.3297 0.2831 0.2493 0.0211 0.0831 0.1441
SCosRec 0.1969 0.3447 0.2930 0.2585 0.0228 0.0886 0.1526
HANN 0.1971 0.3349 0.2971 0.2641 0.0238 0.0924 0.1613

Comparison of model performance on Gowalla

Metric MAP Precision@1 Precision@5 Precision@10 Recall@1 Recall@5 Recall@10
Model
Pop 0.0226 0.0519 0.0359 0.0282 0.0051 0.0272 0.0407
BPR 0.0756 0.1637 0.0983 0.0736 0.0236 0.0747 0.1083
FPMC 0.0762 0.1556 0.0934 0.0694 0.0248 0.0726 0.1065
GRU4rec 0.0582 0.1049 0.0733 0.0781 0.0151 0.0514 0.0831
Caser 0.0922 0.1955 0.1145 0.0575 0.0315 0.0858 0.1228
RCNN 0.0768 0.1767 0.0973 0.0737 0.0269 0.0708 0.1030
CosRec 0.0982 0.2139 0.1187 0.0872 0.0335 0.0878 0.1305
SCosRec 0.1011 0.2192 0.1192 0.0892 0.0343 0.0918 0.1317
HANN 0.1027 0.2199 0.1246 0.0918 0.0367 0.0948 0.1347