Skip to main content

Advertisement

Table 7 Feature contribution experiments for relation extraction

From: A neural joint model for entity and relation extraction from biomedical text

Features ADE BB
  P R F1 P R F1
Word 62.7 69.9 66.1 34.5 20.4 25.6
+dep 63.3 71.0 66.9 42.0 19.9 27.0
+entity 63.4 71.2 67.1 34.1 24.7 28.6
All 67.3 75.7 71.3 42.7 25.2 31.7
  1. Here “+” means only that feature is added.“dep” and “entity” denote dependency type and entity representation features, respectively