Fig. 5
From: A corpus for plant-chemical relationships in the biomedical domain
Example of our corpus units in BioC XML format. The corpus contains information about gold-standard sentences with annotated plant and chemical names along with their IDs, locations of entity names, trigger terms only for positive relationships, and class labels (positives or negatives). The corpus data we provide in our web site are currently divided into three types: (i) 939 primary corpus units from annotation phases 1 and 2; (ii) 68 later-annotation corpus units after harmonizing disagreements; and (iii) 102 rule development corpus units