Journal of Biomedical Semantics

Table 5 Performance of CNN-based models and LSTM-based models in Ablation Study

From: Neural side effect discovery from user credibility and experience-assessed online health discussions

Systems	Components			Evaluation Metrics
	CW	UE	CA	Pre.	Rec.	F₁
LSTM-Vanilla				0.6173	0.407	0.4335
LSTM-WPE	✓			0.6376	0.4344	0.4503
LSTM-WPEU	✓	✓		0.6064	0.5001	0.4896
LSTM-NEAT	✓	✓	✓	0.6197	0.5134	0.5064
CNN-Vanilla				0.7214	0.5503	0.5637
CNN-WPE	✓			0.7423	0.5799	0.5804
CNN-WPEU	✓	✓		0.6923	0.6350	0.5910
CNN-NEAT	✓	✓	✓	0.7066	0.6431	0.6139

In the Components column, CW, UE, CA denote Credibility Weights, User Expertise and Cluster Attention module components, respectively. In the Evaluation Metrics column, Pre., Rec. and F₁ denote Precision, Recall, and F₁ score

Back to article page

ISSN: 2041-1480

Contact us

General enquiries: journalsubmissions@springernature.com