Table 3.

Comparative analysis with unimodal and multimodal approaches

	Modality	Model	Accuracy(%)	Precision(%)	Recall(%)	F1-score(%)
Unimodal	Text	LSTM	77.24	78.36	81.52	79.92
	Text	BiLSTM	79.74	83.32	81.21	82.23
	Text	BERT	83.45	81.67	82.64	82.15
	Graph	GCN	86.42	87.34	85.57	86.44
	Graph	GraphBERT	87.36	86.56	87.79	87.19
Multimodal	Text+Graph	Doc2Vec+Node2Vec	90.17	89.64	88.75	89.19
	Text+Graph	TFIDF+Doc2Vec+Node2Vec	92.34	91.23	90.34	90.74
	Text+Graph	Proposed	95.24	94.67	92.34	93.48