Table 8 Per-entity results for Swedish with word2vec from EHR texts using concatenated word and lemma embeddings

From: Recent advances in Swedish and Spanish medical entity recognition in clinical texts using deep neural approaches

Feature	Algorithm	Entity	Precision	Recall	F1-score
L,W	word2vec	Body part	83.07	94.36	88.36
		Disorder	75.91	73.25	74.55
		Finding	69.97	70.87	70.42
		Avg.	74.64	77.49	76.04

Back to article page