Fig. 4
From: Improving biomedical named entity recognition with syntactic information
Syntactic information extraction. Three types of syntactic information extracted for an example “Dihydropyrimidine dehydrogenase deficiency is an autosomal recessive disease” in the biomedical domain. The context features and their corresponding POS labels, syntactic constituents, and dependency relations for \(x_5\)=“deficiency” are highlighted in part a, b, and c respectively