• 1
• 2
• 3
• 4
• 5
• 6
• 7
39
Text-mining methods used 
for information extraction 
in plant scientific papers
5. Relation extraction
40CC-BY
Relation extraction
41CC-BY
Machine learning
Manual annotation of relations can be used 
to learn methods. The aim is the 
differentiation of instances where there is a 
relation and instances where there is none. 
Different algorithms are popular:
- SVM (support vector machines)
- Artificial neural networks (neural networks)
The majority of the systems used rely on syntactic 
analysis
42CC-BY
Syntactic analysis
Example of a syntactic dependency path
(provided by the Stanford Parser)
LEC2 is implicated in repression of GA biosynthesis.
NNP VBZ VBN NN NNP NNSIN IN
passive nominal 
subject
passive auxiliary case marking
nominal modifier
case marking
compound
nominal modifier