Table 7. The measurement metrics for the entire dataset in the examination recommendation task.

Status	Precision	Recall	F1-score
0	0.98	0.94	0.96
1	0.74	0.89	0.81
AC	0.93
MA	0.86	0.91	0.88
WA	0.94	0.93	0.93

We conducted precision, recall, and F1-score measurements in positive (status =1, examination undertaken) and negative (status =0, examination not undertaken) cases. AC was calculated for the overall dataset. MA precision/recall/F1-score and WA precision/recall/F1-score were calculated through weighted and macro averages of the 10 examination categories. AC, accuracy; MA, macro average; WA, weighted average.