Table 4 Comparison of the AUC and ACC performance (%) of various knowledge tracing methods on the C++, C, Java, and Python datasets. The best results are highlighted in bold, and the second-best results are underlined.
From: Difficulty aware programming knowledge tracing via large language models
Methods | C++_AUC | C++_ACC | C_AUC | C_ACC | Java_AUC | Java_ACC | Python_AUC | Python_ACC |
|---|---|---|---|---|---|---|---|---|
DKT1 | 70.9 | 69.7 | 71.7 | 66.5 | 71.8 | 69.7 | 74.0 | 68.0 |
DKVMN8 | 71.5 | 69.9 | 66.9 | 64.2 | 71.8 | 70.3 | 73.8 | 69.0 |
SAKT53 | 71.9 | 70.6 | 72.5 | 67.0 | 73.0 | 69.9 | 74.5 | 68.8 |
EKT7 | 72.0 | 70.0 | 73.2 | 69.8 | 75.0 | 71.7 | 73.3 | 67.0 |
Code-DKT19 | 71.2 | 69.9 | 71.8 | 67.4 | 71.8 | 67.9 | 72.5 | 69.1 |
DKT+54 | 73.2 | 71.0 | 72.9 | 69.7 | 74.9 | 72.6 | 75.5 | 68.5 |
DIMKT50 | 72.5 | 71.7 | 75.1 | 68.7 | 73.5 | 71.8 | 75.9 | 68.9 |
QIKT55 | 73.4 | 71.6 | 72.4 | 67.3 | 72.2 | 70.8 | 73.8 | 69.5 |
TCKT56 | 73.6 | 70.9 | 73.1 | 68.4 | 72.6 | 71.8 | 76.0 | 67.7 |
THKT57 | 73.8 | 71.8 | 75.2 | 70.0 | 74.5 | 72.5 | 74.9 | 70.0 |
DPKT | 76.2 | 74.5 | 75.7 | 71.5 | 75.4 | 72.9 | 76.8 | 68.1 |
