Skip to main content
Fig. 4 | BMC Bioinformatics

Fig. 4

From: A corpus for plant-chemical relationships in the biomedical domain

Fig. 4

The overall method for extracting plant–chemical relationships from abstracts. To extract plant–chemical relationships from biomedical articles, our model annotates plant and chemical names in the articles using LingPipe and ChemSpot, respectively. Then, the system splits annotated abstracts into sentences using the LingPipe sentence splitter to apply the Stanford Dependency Parser, which provides dependency parse trees for each sentence. Finally, the rule-based model checks the grammar structure of the dependency parse trees and the trigger words in sentences to extract relationships

Back to article page