Mining Precedence Relations among Lecture Videos in MOOCs via Concept Prerequisite Learning

Table 2

Concept prerequisite predicting results.

Methods	Measure	Data mining	Geometry	Physics	Precalculus	AVG
M1	A	83.5	91.9	85.7	89.5	87.7
	P	77.7	88.2	75.2	86.5	81.9
	R	74.3	85.7	63.3	82.0	76.3
	F1	75.8	86.9	68.5	84.1	78.8
	AUC	90.2	96.4	91.5	95.3	93.4
NN	A	74.3	82.1	87.4	90.1	83.5
	P	61.6	69.3	53.5	71.8	64.1
	R	73.8	77.5	74.0	76.9	75.6
	F1	67.0	73.0	61.6	73.6	68.8
	AUC	84.3	90.1	86.2	90.8	87.9
RefD	A	66.2	61.9	74.9	81.9	71.2
	P	51.7	42.4	49.9	75.1	54.8
	R	76.2	62.3	49.6	69.4	64.4
	F1	61.4	50.4	49.4	72.1	58.3
	AUC	69.5	62.4	67.7	79.2	69.7
BERT	A	86.0	92.8	87.4	91.1	89.3
	P	79.7	88.1	75.6	86.5	82.5
	R	81.1	89.3	73.9	87.9	83.1
	F1	80.3	88.6	74.4	87.1	82.6
	AUC	92.2	96.9	92.3	96.6	94.5
GE	A	85.5	91.2	82.4	90.1	87.3
	P	76.5	86.5	74.4	83.4	80.2
	R	81.5	84.1	75.9	81.9	80.9
	F1	78.9	85.3	74.9	82.6	80.4
	AUC	91.7	97.0	93.3	95.4	94.4
Proposed	A	87.0	93.6	88.2	90.9	89.9
	P	83.7	91.3	75.6	86.7	84.3
	R	78.8	88.7	78.0	86.6	83.0
	F1	81.1	89.8	76.7	86.6	83.6
	AUC	92.8	97.4	93.0	96.0	94.8

Bold fonts represent the maximum value of a measure on a dataset.